Exploring what the EU Destruction Ban Changes Inside Your Fashion Operation After 19 July 2026. An intelligence report from the fashion.business team.
1. The write-off code
In most fashion businesses of any size, the decision to destroy stock is not made by anybody who has ever seen the stock.
It is made in a finance review. A season closes, the sell-through report lands, and a quantity of product is identified as having no realistic route to margin. It is written down, usually to zero or to something close enough to zero that the difference stops mattering. The entry appears in the ledger as a stock adjustment. It carries a reason code, and the reason code is almost always a financial one: obsolete, damaged, aged, provision. Nobody records what the garments were, how much they weighed, or what happened to them next.
What happened next was handled by an operations manager. Not a sales manager, because by that point the product had left the commercial conversation entirely. Somebody arranged a collection. A waste contractor took a number of pallets, issued a transfer note describing the load as mixed textiles, and the matter closed. In a well-run business the whole sequence took under a fortnight and cost less than the storage would have.
From 19 July 2026, that sequence is unlawful for large companies placing apparel, clothing accessories or footwear on the EU market. Not the write-down, which remains a perfectly ordinary accounting judgement. The disposal. The prohibition in Article 25 of the Ecodesign for Sustainable Products Regulation removes destruction as a default option, and the delegated regulation that accompanies it removes the informality that made the old sequence so cheap.
Here is the part that has been under-reported. The regulation does not simply ask brands to stop destroying stock. It asks them to prove what they did instead, at a level of granularity that the stock adjustment entry was never designed to carry. Where destruction does happen under one of the permitted exceptions, the brand must hold documentary evidence of the specific ground relied on, retain it for five years in electronic form, produce it to a national authority within 30 days of a request, and hand a statement of the applicable derogation to the waste treatment operator receiving the goods. Where destruction does not happen, the brand must still publish, annually, the number of units it discarded, their weight, the reason, and the proportion that went to reuse, recycling, other recovery or disposal.
The finance system that generated the write-off knows the value. It rarely knows the weight. The warehouse knows the pallets. It rarely knows the reason. The 3PL knows the collection date. It almost never knows which of ten legal grounds justified it. The compliance obligation sits across all three, and in most brands there is no single record that joins them.
That gap is the subject of this report. The ban itself is straightforward and, for most brands, not commercially painful. The evidence trail underneath it is neither, and it is the thing that will be tested first.
2. The landscape: how destruction became routine, and what closed it
The economics that made incineration rational
Destroying new clothing is not, in the ordinary case, an act of corporate vandalism. It is the endpoint of a set of incentives that made every other option worse.
Consider a garment that cost £9 landed and was priced at £45. It fails at full price, fails at 30% off, fails in outlet, and is still sitting in a national distribution centre eleven months later. Its carrying cost is real: warehouse space, insurance, stock counting, system records, the working capital tied up in a provision that the auditor keeps asking about. Its residual value is close to nothing. Selling it into a jobber market risks it reappearing at £6 in a market where the brand sells at £45. Donating it means finding a charity willing to take 4,000 units of one style in sizes that skew large, then paying to transport them. Recycling it means paying a gate fee for fibre that mixed-composition garments make hard to process.
Against that, a waste contractor will remove the problem for a known price on a known date, and the accounting treatment is clean. For thirty years, that arithmetic held. The European Environment Agency, in a briefing published in March 2024 on the destruction of returned and unsold textiles, estimated that between 4% and 9% of all textile products placed on the European market are destroyed before they are ever used, which it put at between 264,000 and 594,000 tonnes each year. The Commission attaches roughly 5.6 million tonnes of CO2 equivalent to that practice annually, a figure it compares to the total net emissions of Sweden in 2021.
Two structural changes made those numbers grow rather than shrink. The first was the collapse of the return as an exception. E-commerce return rates in apparel run between 20% and 50% depending on category and market, and a meaningful proportion of returned goods never re-enter saleable stock, because the cost of inspecting, cleaning, repackaging and restocking a single unit exceeds its remaining margin. The second was the shortening of the range cycle, which multiplied the number of options carried and reduced the sell-through window for each.
The result was a category of stock that was physically new, commercially dead, and administratively invisible.
France moved first
The European measure did not appear from nowhere. France legislated against it in the Anti-Waste for a Circular Economy law of 10 February 2020, which prohibited the destruction of unsold non-food products and took effect from January 2022. Estimates of the volume it captured vary. The figure the Commission and European press have most often cited is around €630 million of unsold products destroyed annually in France before the prohibition, alongside an estimate of some 20 million returned items discarded each year in Germany.
The French law demonstrated two things. It proved that a national prohibition was enforceable in principle. It also demonstrated that a national prohibition inside a single market simply relocates the problem, because stock is mobile and disposal routes are cross-border. That argument, more than any other, is why the measure moved to EU level.
What the ESPR actually is
The Ecodesign for Sustainable Products Regulation, Regulation (EU) 2024/1781, was published in the Official Journal on 28 June 2024 and entered into force twenty days later, in July 2024. It replaces the 2009 Ecodesign Directive and extends the ecodesign concept well beyond energy-related products to almost all physical goods placed on the EU market.
Most of the ESPR is a framework. It empowers the Commission to set product-specific requirements on durability, reparability, recycled content, substances of concern and the Digital Product Passport, delivered through delegated acts over the coming decade. Textiles and footwear were confirmed as a priority group in the first working plan.
Three provisions, however, are self-executing and do not wait for product-specific rules:
Article 23 places a prevention duty on all economic operators, regardless of size, to take the measures that can reasonably be expected of them to avoid the need to destroy unsold consumer products in the first place. There is no small-company exemption from this duty. It is frequently overlooked in commentary aimed at smaller brands, who read the headline size thresholds and conclude the file is not theirs.
Article 24 requires large enterprises that discard unsold consumer products, whether directly or through a third party, to disclose publicly the number of units, their weight, the reasons, the waste treatment operations applied, and the preventive measures taken. This obligation has been running since the first full financial year after entry into force.
Article 25 prohibits the destruction of unsold consumer products listed in Annex VII. Annex VII currently contains apparel, clothing accessories and footwear, identified by customs classification: chapters 61 and 62 for knitted and non-knitted apparel and clothing accessories, heading 4203 for leather apparel and accessories, and headings 6401 to 6405 for footwear. The prohibition applies to large enterprises from 19 July 2026 and to medium-sized enterprises from 19 July 2030. Micro and small enterprises are exempt from the prohibition, though not from the Article 23 prevention duty.
On 9 February 2026 the Commission adopted the two instruments that made Article 25 operable. Commission Delegated Regulation (EU) 2026/296 defines ten derogations from the prohibition in its Article 2, and took effect on 12 May 2026. Commission Implementing Regulation (EU) 2026/2 sets the standardised disclosure format, the product categorisation and the verification framework; it entered into force on 2 March 2026 and applies from 2 March 2027. The Commission confirmed the entry into application of the prohibition in a statement on 17 July 2026, three days before this report was published.
The definition that catches people out
There is a trap in the word destruction, and it has caught out several brands who believed themselves compliant.
Article 2(34) of the ESPR defines destruction as the intentional damaging or discarding of a product as waste. Read quickly, that sounds like it describes a bonfire. It does not. Discarding as waste is the operative limb, and it captures consignment to a waste treatment operation of any kind.
That includes recycling, and three things establish the point rather than one. Recital 3 of the Delegated Regulation lists what does not constitute destruction, and the list is limited to donating, remanufacturing, refurbishing and preparing for reuse. Recycling is not on it. The Annex I disclosure template in the Implementing Regulation treats preparing for reuse as its own column and groups recycling, other recovery and disposal together under destruction. And the European Environment Agency, in the briefing this report cites for its volume estimates, treats recycling as destruction on the same reasoning, because a product must be shredded or dismantled to be recycled and therefore ceases to exist as a product.
So a brand that responds to the ban by diverting its end-of-season surplus into fibre-to-fibre recycling has not complied with Article 25. It has destroyed the goods within the meaning of the regulation, and it needs a derogation to have done so lawfully. Note that this is a conclusion drawn from the exclusion list and the disclosure architecture rather than a phrase lifted from the definition itself, and that some published summaries render Article 2(34) loosely enough to obscure it. It is nonetheless the reading every major firm briefing has taken.
What does not constitute destruction is donation, remanufacturing, refurbishment, and preparation for reuse. The hierarchy the regulation imposes is therefore sharper than the familiar waste hierarchy. Keeping the product in use as a product comes first. Recovering the material is already a form of failure in the eyes of the regulation, permitted only when one of the ten grounds applies. Even then, where destruction is lawful, it must follow the waste hierarchy, with recycling prioritised over other recovery and over disposal.
For an operations director who has spent two years building a textile recycling partnership as the responsible answer to surplus, this is an uncomfortable reversal. The partnership is not wasted. It is simply no longer the first answer, and using it as the first answer is now a breach.
The measurement problem nobody has solved
There is a further point buried in the disclosure regime that deserves more attention than it has received.
The EEA range of 4% to 9% is wide because nobody knows the denominator. Customs data records what is imported into the Union with reasonable precision. It records nothing about what is manufactured inside the Union and sold there, which is a material share of the market in footwear, in Italian and Portuguese apparel manufacture, and in the growing near-shored segment. There is no register of units placed on the EU market. There is only a set of estimates built from trade statistics and survey work.
The Article 24 disclosure regime is the first serious attempt to fix the numerator: what was discarded, by whom, in what quantity, for what reason. It leaves the denominator untouched. Over the coming decade the Digital Product Passport is the mechanism that could close that gap, because unit-level identification, of the kind that infrastructure such as id.codes is being built to support, would let a regulator establish how many units of a category were placed on the market irrespective of whether they crossed a customs border or were made in Porto. Until something of that sort exists, the EU will be enforcing a percentage reduction against a total it cannot measure, and brands will be publishing absolute numbers with no context in which to read them.
That is not a reason to delay compliance. It is a reason to expect the reporting requirements to become more granular, not less, as the infrastructure catches up.
3. Timeline overview
10 February 2020. France adopts the Anti-Waste for a Circular Economy law, prohibiting the destruction of unsold non-food products. Effective from January 2022. The template for the EU measure.
December 2023. Parliament and Council reach political agreement on the ESPR, including the destruction ban for textiles and footwear.
28 June 2024. Regulation (EU) 2024/1781 published in the Official Journal.
July 2024. ESPR enters into force. The Article 23 prevention duty applies immediately, to operators of every size. The Article 24 disclosure obligation begins to run for large enterprises from their first full financial year after this date.
16 October 2025. The revised Waste Framework Directive, Directive (EU) 2025/1892, enters into force, requiring every member state to establish mandatory extended producer responsibility schemes for textiles and footwear. This is the financial mechanism that will sit alongside the destruction ban.
9 February 2026. The Commission adopts Delegated Regulation (EU) 2026/296 (ten derogations) and Implementing Regulation (EU) 2026/2 (disclosure format, product categories, verification framework).
2 March 2026. The Implementing Regulation enters into force.
12 May 2026. The Delegated Regulation setting out the ten derogations takes effect.
19 July 2026. The Article 25 prohibition applies to large enterprises for apparel, clothing accessories and footwear. The derogations in the Delegated Regulation apply from the same date.
2 March 2027. The Implementing Regulation applies. The prescribed Annex I format governs financial years starting on or after this date. This is later than it sounds and is widely misreported. A brand with a March year end has its first affected financial year running from March 2027 to March 2028, and its first disclosure in the prescribed format is due by March 2029. The Commission's own summary language describes the format as applying from February 2027, which has been repeated across the trade press without the financial-year qualifier.
19 July 2027. Under Article 25, the Commission must publish consolidated information on the destruction of unsold consumer products on its website, and again every 36 months thereafter. This is the first Union-level picture of the sector's discard behaviour, and it arrives roughly two years before most brands publish their first standardised individual disclosure.
17 April 2028. Textile extended producer responsibility fees apply at the latest, under Directive (EU) 2025/1892. Producers pay a fee for each textile product placed on the market. Used textiles assessed as fit for reuse are outside the fee.
2028. Mid-term review point for the ESPR working plan. The most likely moment for the Annex VII product scope to be extended beyond apparel and footwear.
19 July 2030. The prohibition and the disclosure obligation extend to medium-sized enterprises. Micro and small enterprises remain exempt from both, but not from the Article 23 prevention duty.
2 March 2031. The Commission must review the Implementing Regulation, including the relevance of the Annex I format and the Annex II product delimitation, and present the results with a draft revision if appropriate. Worth knowing before commissioning bespoke engineering against the current template.
4. The central argument
The destruction ban is not difficult to obey. It is difficult to evidence. Every serious cost, risk and system implication in this file flows from that distinction, and brands that read the measure as an environmental policy rather than a records policy will be the ones caught out.
4.1 Who is actually in scope, and why the answer is contested
Start with the question every brand asks first, because the professional commentary does not agree on the answer.
The ESPR allocates obligations by enterprise size. Most authoritative commentary, including analysis that works directly from the regulation, applies the classification in Directive 2013/34/EU, the Accounting Directive. On that test a large undertaking exceeds at least two of three thresholds: more than 250 employees, more than €50 million in net turnover, and more than €25 million in balance sheet total. Other commentary applies Commission Recommendation 2003/361/EC, the SME definition, under which an enterprise ceases to be medium-sized above 250 employees combined with either turnover above €50 million or a balance sheet above €43 million. At least one legal blog has asserted a threshold of 500 employees and €400 million turnover, which appears to import criteria from a different instrument entirely and should be treated with caution.
The disagreement is now resolvable, and it resolves against the majority commentary. The ESPR's own definitions article does not define large enterprise directly. It imports the definitions of SMEs, small enterprises and microenterprises from Article 2(1), (2) and (3) of Annex I to Commission Recommendation 2003/361/EC, and derives everything above them by subtraction. Freshfields reads the cross-reference the same way. So the Recommendation governs, not the Accounting Directive.
That distinction is not academic, because the two instruments treat headcount in opposite ways. Under the Recommendation, headcount is a hard gate: below 250 persons you cannot be large, whatever your turnover. Under the Accounting Directive, headcount is one of three criteria and any two suffice, so a high-turnover business with a small payroll would be large on the financial tests alone. If you have been told the Accounting Directive applies, the source is working from a reasonable inference rather than from the text.
For a brand between £20 million and £100 million, the practical guidance is simpler than the disagreement suggests. Headcount is the dominant trigger. If you employ fewer than 250 people, you are almost certainly medium-sized or smaller under either test, and the prohibition reaches you in July 2030 rather than July 2026. If you employ more than 250 and turn over more than €50 million, you are in scope now under either test. Where the disagreement bites hardest is the business with high turnover and low headcount, which section 4.1a takes up, because there the choice of test decides whether the ban applies today or in 2030.
Three complications matter more than the threshold arithmetic:
Group consolidation. Size is assessed at the level of the undertaking including linked enterprises. A brand employing 40 people, wholly owned by a group that is large, is treated as large. This catches acquired brands, private-equity-held platform companies, and the European subsidiaries of substantial non-EU parents. If your brand was acquired in the last three years and you have not revisited your classification, do that before anything else in this report.
The two-year rule. Under the SME classification, status changes only when thresholds are exceeded or fallen below for two consecutive accounting periods. A single year of exceptional trading does not move you. Nor does a single bad year move you back.
Establishment is irrelevant. The obligations attach to products placed on the EU market, not to companies established in the EU. A UK, US or Turkish brand shipping into the Union is an economic operator for these purposes. Post-Brexit this is the single most misunderstood point among British brands, many of which read EU environmental legislation as somebody else's compliance file. If you sell into the EU through a subsidiary, a distributor, a marketplace or your own direct e-commerce channel, and you are large, Article 25 applies to the goods you placed there.
One further asymmetry is worth naming. Only around a fifth of apparel sold in the EU is manufactured within it. The prohibition therefore lands overwhelmingly on goods made elsewhere and sold into the bloc, which is precisely why it was drafted against market placement rather than establishment. A brand that manufactures in Bangladesh, holds stock in a Dutch third-party warehouse and sells across seven member states is squarely inside the regime, wherever its head office sits.
4.1a Why store estates and pure-plays land on opposite sides of the line
There is a consequence of the classification disagreement that has gone almost entirely unremarked, and it falls hardest on precisely the businesses this report is written for.
The two candidate tests treat headcount in fundamentally different ways.
Under the Accounting Directive, headcount is one criterion among three, and you are large if you exceed any two. A pure-play e-commerce brand turning over €70 million with €30 million of assets and 60 staff is large on the two financial criteria alone. Its headcount is irrelevant to the outcome.
Under the SME Recommendation, headcount is a gate rather than a criterion. An enterprise with fewer than 250 staff cannot be large, whatever its turnover. On that reading, a €150 million online business with 90 employees remains medium-sized until 19 July 2030.
Now apply that to how fashion businesses are actually staffed. A retailer with 40 stores carries hundreds of people on the payroll before a single unit of stock is sold, because selling floor space requires staff on the floor. A wholesale brand of identical turnover employs a fraction of that number, because its route to market is somebody else's store estate. An online pure-play may employ fewer people still, because fulfilment, warehousing and returns processing sit with a third-party logistics provider, and headcount that would once have been internal now appears on the 3PL's payroll rather than the brand's.
The result is that two brands with the same EU revenue, the same product, and materially the same volume of unsold stock can sit four years apart in their compliance obligations, determined by their channel model rather than by anything the regulation is actually trying to influence.
Two refinements soften the picture without removing the asymmetry.
The first is that store headcount is smaller in the calculation than it looks on the payroll. The Recommendation counts headcount in annual work units rather than in bodies. Part-time staff count in proportion to their hours and seasonal workers are averaged across the financial year, so part-time and seasonal staff count pro rata rather than as whole heads. A 40-store estate with 300 names on the payroll may be closer to 170 annual work units once part-time contracts and Christmas temporary staff are converted. Retailers who assume they are obviously over the threshold should run the annual work unit figure before concluding anything, because a good number will find they are not. The Accounting Directive uses an average number of employees during the financial year, defined at member state level, which is calculated differently and is generally less forgiving.
The second is group consolidation, which overrides the whole discussion. The Recommendation's annex distinguishes partner enterprises, held between 25% and 50%, from linked enterprises, held above 50%. Partner holdings are aggregated proportionally; linked enterprises are aggregated in full. A lean online brand majority owned by a private equity platform or by a larger group is therefore assessed together with that group and is large today, regardless of how few people it employs directly. In practice this catches a substantial share of pure-plays above the £50 million mark, because that is the size at which they tend to have taken institutional money.
Neither refinement appears in the ESPR itself. Both come from the Recommendation and its accompanying user guide, which is worth knowing, because the distinction between what the instrument requires and what a guidance document supplies matters if the point is ever argued.
What remains after both refinements is a genuine distortion inside a measure explicitly justified as levelling the playing field. And it runs the wrong way. The pure-play with a 30% return rate is generating more non-resellable stock per pound of revenue than the store operator whose customers try garments on before buying and return a fraction as many. The business model producing the most unsold and unsaleable product is the one most likely to sit outside the ban until 2030, on the strength of having outsourced the people who handle it.
That points at the harder question underneath this regulation, which is returns rather than surplus. Overproduction is visible, seasonal and easy to legislate against. Returns are continuous, invisible in the ledger, and generate non-resellable units one at a time in a warehouse operated by somebody else. The destruction ban addresses what happens at the end of that process. It does nothing about the volume entering it. That is a subject for a separate report.
For which instrument governs the size test, why headcount is the gate, and what annual work units do to the crossing point, see Who is the economic operator?
4.2 The ten derogations, and what each one costs you in evidence
Delegated Regulation (EU) 2026/296 defines, in its Article 2, ten circumstances in which destruction of Annex VII goods remains lawful. Read as a list they look generous. Read as evidence requirements they are considerably tighter, because each ground carries its own documentary burden and the burden falls on the operator.
The grounds, in the order the regulation sets them out:
(a) Dangerous product. The item is dangerous within the meaning of the General Product Safety Regulation (EU) 2023/988 and poses a risk to health or safety that cannot be mitigated by other means. Evidence: a safety assessment under GPSR Articles 6 to 8, or a test report showing non-compliant chemistry with a reference to the applicable law.
(b) Non-compliance with law, other than safety. The product is unfit for purpose because it does not comply with EU or national law for a non-safety reason, including ethical grounds such as forced labour in the supply chain, and destruction is either required by law or is the appropriate corrective action. Evidence: a self-assessment statement identifying the type of non-compliance and the law engaged.
(c) Intellectual property infringement. Established by a final judicial decision, an alternative dispute resolution outcome, a notification from a rights holder or competent authority, or a substantiated internal investigation. Evidence: the decision, notification or investigation record.
(d) Expired intellectual property licence. The product is subject to a valid licence restricting sale after a specified date and that date has passed. Evidence: the licence or contract containing the restriction, plus a justification that destruction is appropriate and proportionate. Note the second limb. Holding the licence is not sufficient; you must show why destruction rather than another route was the proportionate response.
(e) Unsuitable for reuse. The product cannot be prepared for reuse or remanufactured because it is technically unfeasible to remove or permanently obscure labels, logos or design characteristics that are protected by intellectual property or considered inappropriate, the example given being product perpetuating discrimination or stereotypes. Evidence: an inspection report demonstrating that technical options were assessed and found unfeasible, supported by visual evidence, technical analysis or expert opinion.
(f) Damaged product. Physically damaged, deteriorated or contaminated during handling, storage, transport, retail or consumer return, to the point of being unacceptable for consumer use, where repair or refurbishment is not technically feasible or not cost-effective. Recital 9 of the Delegated Regulation expressly contemplates this ground covering products returned by consumers under the statutory right of withdrawal or during a longer trader withdrawal period, which makes the most common case in this industry an expressly anticipated one rather than an inference. Evidence: either documented quality assessment procedures including sorting that prioritises restocking and repair, or an inspection record documenting the damage and the unfeasibility of correction.
(g) Design or manufacturing defect. The product is non-functional because of a design or manufacturing defect and repair is not technically feasible. Evidence: as for (f).
(h) Not accepted for donation. The product was offered to at least three social economy entities within the EU, or listed publicly on the operator's website for a minimum of eight weeks, and was not accepted. Evidence: proof of the offer. This ground is available only where grounds (a) to (g) do not apply.
(i) Social economy entity with no recipient. The goods were received as a donation by a social economy entity, which could find no recipient. Evidence: a declaration to that effect.
(j) Prepared for reuse with no recipient. The goods were prepared for reuse by a waste treatment operator under the Waste Framework Directive and no recipient was found. Evidence: documentation of receipt from the operator and of the failed placement.
Two definitions inside the delegated regulation do a great deal of work.
The first is cost-effectiveness, which governs ground (f) and is where most volume will sit in practice. The regulation defines it as the cost of repairing the product not outweighing the total cost of destruction plus replacement, where replacement is measured across materials, manufacturing, packaging, transport, stocking and administration. The comparator is replacement, not resale value. Read that carefully, because it cuts the opposite way to how brands assume. The comparator is not the residual value of the garment. It is what it would cost you to make and land another one. A £9 landed cost garment with a £2 repair is cost-effective to repair on this definition, even if the garment will never sell at full price. Brands accustomed to justifying disposal by reference to the item's market value are using the wrong denominator and will produce evidence files that do not support the ground they claim.
A reader who checks the parent regulation will look for an eleventh ground and not find it. Article 25(5) of the ESPR contemplates destruction being permitted where it is the option with the least negative environmental impact, but the Commission did not carry that ground into the delegated act, reasoning that reuse of Annex VII products is always environmentally preferable to recycling. There is therefore no environmental-benefit derogation, and an argument that incineration was the greener option is not available.
One serious criticism of this definition deserves an answer rather than silence. COSH!, the Belgian sustainable fashion platform, has argued through its founder Niki de Schryver that if cost-effectiveness comes to be read in practice as cheaper than repairing the item right now, and if extended producer responsibility schemes continue to subsidise recycling, then repair loses on price for structural rather than technical reasons, and the derogation swallows the rule.
The first half of that is answerable on the text. The definition in Article 1 of the Delegated Regulation does not support the narrow reading, because the comparator is replacement cost and not the cost of doing nothing. The second half is not answerable at all, and this report will not pretend otherwise. Whether national authorities hold that line, against an EPR regime pushing economics in the opposite direction, is genuinely open, and it will be settled by enforcement practice rather than by drafting.
The second definition doing heavy work is the social economy entity, defined by reference to Article 3(4i) of the Waste Framework Directive. This matters because ground (h) is not satisfied by offering stock to any charity. It requires offers to entities meeting that definition, or a public listing running the full eight weeks. Eight weeks is a long time in a warehouse cycle, and it is a period that has to start before you need the space, not after. A brand that decides in week one of a clearance push that it will donate what does not sell has already lost the ability to rely on ground (h) within the same season.
Three practical constraints apply across all ten grounds. Documentation must be held in electronic form and produced to a competent authority within 30 days of request, and it is retained on two different clocks that are easy to conflate. Derogation evidence runs for five years from the destruction, under Article 3 of the Delegated Regulation. Documentation demonstrating the delivery and reception of the discarded goods, including the waste treatment operator's statement, runs for five years from the date the disclosure was made, a rule stated consistently by Cooley, Bird & Bird and Beveridge & Diamond in their 2026 briefings. The disclosure clock is normally the later of the two, so a single five-year archive dated from the collection will fall short at the back end. It may be prepared collectively where the same circumstances affect multiple products, which is the provision that makes batch-level compliance workable rather than requiring a file per garment. And under Article 4, the operator must give the receiving waste treatment operator a statement identifying the derogation relied upon. That last obligation is the one most likely to be missed, because it requires the compliance judgement to be communicated outward, to a supplier who has never previously needed it, at the moment of collection.
The ten grounds are set out as a working table, with the evidence artefact and owning function against each, in The ten derogations: a working reference. The cost-effectiveness sum is run against a real consignment in One write-off, start to finish, which shows why the comparator disqualifies most of what brands assume it permits.
4.3 What you must publish, and the tolerance you are allowed
Implementing Regulation (EU) 2026/2 does something the ban itself does not. It makes the numbers public and comparable.
Before the disclosure detail, one scope point belongs here rather than with the ban, because it is the second of the two tests that decide whether a unit is caught at all. Article 25(1) reaches Annex VII products that have been placed on the Union market and are primarily intended for consumers. Components, intermediate products and goods not primarily intended for consumers fall outside both the prohibition and the disclosure. For a brand selling the same fabric or trim into both finished product and trade channels, the classification of a given consignment is not automatic.
The disclosure covers a far wider product scope than the ban. While Article 25 currently reaches only Annex VII apparel and footwear, the disclosure format lists over 50 product categories by customs code, spanning electronics, appliances, furniture, toys, tyres, hygiene products and home textiles. For a fashion business with a homeware line, a small-electricals accessory range or a licensed toy collaboration, the reporting scope is broader than the destruction scope, and the two should not be scoped as a single project.
Annex I of the Implementing Regulation prescribes the fields. In summary:
- Entity identification. Legal name, EUID, whether the disclosure is standalone or consolidated, and the financial year covered.
- Product data. Customs code, units discarded, weight in kilograms, whether packaging is included, and the reason for discarding.
- Waste treatment outcome. The percentage split across reuse, recycling, other recovery, disposal and unknown.
- Prevention measures. Both those taken in the preceding financial year and those planned.
Products are reported at two-digit customs code level by default, with four-digit granularity required for categories listed in Annex II. Components, intermediate products and goods not primarily intended for consumers are excluded.
The disclosure must be published within twelve months of the financial year end, on the company website, or within a CSRD sustainability report with a link from the website.
The single most useful number in the entire secondary legislation is the tolerance. Annex III of the Implementing Regulation establishes a risk-based verification framework, and it treats a discrepancy of less than 10% between the figures disclosed and the supporting documentation as compliant. Above 10%, an authority may make a finding of non-compliance.
That is a workable margin, and it is also a demanding one for a brand whose only record of discarded stock is a value-based ledger entry. It means your systems need to be accurate to within one part in ten on unit counts and on weight, across all EU operations, including stock disposed of by a third party on your behalf.
Annex III also sets out what will attract attention. The absence of a disclosure. Figures that look implausibly low for the size and activity profile of the operator. A high proportion of the waste treatment split reported as unknown. A history of previous non-compliance. And cross-referencing against other data sources, tax declarations among them.
Read that list from the perspective of a market surveillance officer with limited resources and a sector to police. The cheapest possible enforcement action is to compare a brand's published discard figures against its published financial statements and its competitors' disclosures, and to open a file wherever the number looks too good. The first enforcement wave will not be triggered by inspections at warehouses. It will be triggered by a spreadsheet.
The three obligations are set side by side, including the placing precondition, in The ten derogations: a working reference.
4.4 The system gap: where write-offs actually live
Here is where the regulation meets the reality of how fashion businesses are run, and where most of the work will fall.
Non-resellable stock is typically recorded as a stock adjustment at the point it is written down to zero or near zero. It is historically a finance decision, taken at the level of value rather than unit, and recorded in the general ledger or the ERP stock module in value terms. Once the decision is made, an operations manager arranges disposal by whatever means has been chosen. The sales organisation has no further involvement, because the product has left the commercial pipeline.
That process produces exactly one of the four data points the regulation requires. It produces a quantity, sometimes. It rarely produces a weight, almost never produces a legally framed reason, and effectively never produces a verified waste treatment outcome, because the outcome is determined by the contractor after collection and is reported back, if at all, as a tonnage on a waste transfer note covering a mixed load.
The weight problem is the most tractable, and the good news is better than most brands expect. Product weight already exists in the majority of businesses of this size, in one of two places and usually both. The PLM system holds it because weight is captured during development for costing, packing and specification purposes. The ERP holds it because weight is required for customs declarations on import and for shipping rate calculation on outbound e-commerce. A business that imports and ships parcels has already solved the weight problem for other reasons. Which means that a value-based write-off, expressed in units and joined to an existing product master weight, produces a compliant weight figure by multiplication. No new data capture is needed. What is needed is the join, and the discipline of writing off in units as well as in value.
The reason problem is harder, because the reason codes that exist are financial and the reasons the regulation recognises are legal. An obsolete code tells you why the item had no value. It does not tell you whether the item was damaged beyond cost-effective repair, offered unsuccessfully for donation, or bound by an expired licence. Those are different derogations with different evidence packs, and a brand that maps every write-off to a single generic ground will produce a disclosure that fails the plausibility test in Annex III.
The outcome problem is the one most brands have no visibility of at all, because the answer is held by a third party. If your 3PL arranges disposal, your 3PL knows what happened. Your contract with them almost certainly does not require them to tell you, and their waste contractor's transfer note describes a load, not your units. The regulation does not care that the information sits with a supplier. Article 24 applies to products discarded directly or through a third party.
There is a reason for this that is architectural rather than negligent, and it is worth stating plainly because most brands assume the gap is a symptom of their own disorganisation.
When a fashion business specifies a PLM or an ERP, the scope of the project runs from concept through development, sourcing, costing and production, and stops at the point the goods are received and available to sell. End of useful life, and the handling of excess stock, are almost never considered by the company adopting the technology. Ben Muis has spent more than 35 years in the fashion industry, and in the 17 years since moving into consultancy has been involved in over 100 PLM projects. Before that he worked inside the industry itself, for several US and European companies, where he held operational responsibility for the brands, with the full product lifecycle under his control. That is to say he made these write-off decisions before he spent seventeen years watching other people make them. He puts the consequence directly: a write-off of stock is treated as something that happens in the dark, a byline on the financials rather than a predetermined part of the lifecycle of a product stock unit. The product either sells as useful or it does not, and the second outcome is not the goal, so it is not the route anybody designs.
Returns are the instructive exception. Where returns carry significant value, brands do invest serious time and energy in mapping the process systematically, because the money is visible and recurring. The result is an odd asymmetry inside the same business: a well-documented returns workflow feeding into a disposal decision that nobody has ever mapped at all. The regulation now attaches to the undesigned half.
The practical shape of the fix is therefore a joining exercise rather than a systems replacement:
- Write off in units, not only in value, at the point of the finance decision.
- Extend the reason code list from financial categories to the ten legal grounds, with the financial reason retained alongside.
- Join the unit write-off to the product master weight already held in PLM or ERP.
- Amend third-party logistics and waste contracts to require the treatment outcome to be reported back per consignment, mapped to your write-off reference.
- Hold the evidence pack for each derogation claim against the same reference, electronically, for five years.
None of these five is a large project on its own. Taken together, and started in the wrong quarter, they become one.
What a compliant record looks like end to end is followed through in One write-off, start to finish. Why a third-party logistics provider discarding on your instruction does not take the obligation on is covered in Who is the economic operator?.
4.4a The architecture question underneath all of this
The unsold goods regime is one instrument. The Digital Product Passport, textile extended producer responsibility, deforestation due diligence on leather and the substances of concern regime are all arriving on the same product record within the same decade, and each of them asks a version of the same question: what is this unit, where did its materials come from, and what happened to it.
That makes the architecture decision more consequential than any individual compliance project, and it is being made in most businesses this year by default. Three principles are worth holding onto, drawn from the experience of specifying these systems rather than from the regulations themselves.
Cascading, not gated. The instinct among some technology providers has been to run traceability as a managed service, with the vendor sitting between the brand and the supply chain as the party that gathers and verifies. Companies running traceability as a service are, in the assessment of practitioners who have advised the vendors building these tools, already on an unviable path. There is no office of people large enough to track and trace each unit and the origins of its components across global apparel manufacture. It has to be a technology solution that is open for vendors at any level of the supply chain to use, with data trickling down and pushing up through the tiers rather than being collected centrally by a service provider. Where a brand selects a tool that gates access in order to protect a service revenue, it has bought something that cannot scale to the problem it was bought for.
Reuse over resubmission. Efficiency of data flow matters as much as data capture. Suppliers cannot be expected to hire teams to do compliance administration. A factory working with 15 or 20 brands, each with its own supply base beneath it, does not have full traceability in place and will not build it separately for each customer. Whatever technology a brand puts in place on its own side has to be usable not only by the brand but by its suppliers, and it has to let them confirm existing information rather than submit from scratch every time. Data should be reused wherever it remains accurate. A system that asks the same factory the same question twelve times a year in twelve different formats will be answered badly, and badly answered data is what a market surveillance authority will eventually find.
The proof is heavier than the data. This is the point most often missed in sustainability data programmes, and it is the one that will bite hardest in this particular regime. Structured data is not, in itself, the problem. A line of data describing a unit, its weight, its reason code and its destination is small, and remains small even when it is held in several places, provided it is not replicated infinitely across multiple databases in multiple jurisdictions. Evidence is a different order of magnitude. The moment a regulation requires you to prove rather than to state, you are no longer storing lines. You are storing PDF documents, inspection photographs, test reports, safety assessments and scans, and those files are very often not optimised for upload. Attach proof to everything, as the derogation regime in this report requires and as the Digital Product Passport will require across a far wider set of claims, and the data effort multiplies rather than increments.
The architectural consequence follows directly. If a DPP system does not tap directly into the source where documents are already stored, those documents will be duplicated somewhere else by definition, and the effort is enlarged again. The same applies to every parallel compliance regime that wants sight of the same certificate. The purpose of all this gathering is partly to create visibility of what is actually happening, which in turn informs refinement of the legislation itself. But an approach that copies every proof file into every system that might one day want to look at it ends up counteracting the initial goal. The discipline is to store the evidence once, at source, and to point at it from wherever it is needed.
Applied to the immediate problem in this report, those three principles point at the same conclusion the five steps above arrive at from the operational direction. Do not build a separate unsold goods compliance system. Extend the record you already hold, join it to the systems that already know the weight and the quantity, and require the outward flow of information from your logistics and waste partners rather than rebuilding what they already know inside your own four walls.
4.5 What enforcement will look like, and one myth to discard
Enforcement is national. Article 74 of the ESPR requires member states to lay down penalties that are effective, proportionate and dissuasive, and to enforce them through their own market surveillance authorities. The regulation does not set an EU-wide financial ceiling or floor.
This is worth stating plainly because a claim has circulated in professional commentary that ESPR non-compliance carries fines of up to 4% of EU turnover. That figure belongs to the EU Deforestation Regulation, which does set a minimum maximum penalty of 4% of Union-wide turnover in its Article 25. The ESPR does not. Brands scoping their exposure on the basis of a 4% headline number are working from a misattribution.
What the ESPR does provide, and what matters more commercially than the size of any fine, is the market access consequence. The safeguard and market surveillance machinery of the regulation permits authorities to restrict or prohibit the making available of a non-compliant product on the market. For a brand whose EU business depends on continuity of supply into a distribution network, an intervention of that kind is materially worse than a penalty.
The regulation also anticipates the obvious avoidance route, and it does so twice, through two mechanisms that are frequently described as one. Article 25(2) is a standing prohibition already in force: an economic operator that is not itself subject to the ban may not destroy unsold consumer products supplied to it for the purpose of circumventing that prohibition. Article 25(5) is something different, a reserve power allowing the Commission to extend the ban or the Article 24 disclosure to micro and small enterprises by delegated act, but only where there is sufficient evidence that such enterprises are being used to circumvent them. One bites today on a specific intent; the other requires a pattern to become visible before it can be used at all.
Expect enforcement to be uneven across member states in the first two years, weighted towards the jurisdictions with existing textile enforcement infrastructure, France and the Netherlands prominent among them. Expect it to intensify once comparison becomes possible, which happens earlier at Union level than at brand level. See section 8.
5. On the record
The positions taken publicly around this measure tell you where the pressure will come from next. Four are worth understanding.
The Commission: waste as a competitiveness argument
The framing the Commission has chosen is notable for what it does not emphasise. Jessika Roswall, Commissioner for Environment, Water Resilience and a Competitive Circular Economy, introduced the February 2026 measures with the observation that "The numbers on waste show the need to act", and then argued the case in economic rather than purely environmental terms: that the measures would help the textile sector move towards circular practice, improve competitiveness, and reduce dependencies. The Commission's own summary of the measures leads on cutting waste, reducing environmental damage, and creating a level playing field for companies already operating sustainable models.
That last phrase is the operative one for anybody trying to predict enforcement appetite. The measure is being sold internally as a fairness instrument: brands that already carry the cost of resale, repair and donation infrastructure have been competing against brands that carry no such cost because they incinerate. A regulator who believes it is levelling a playing field enforces differently from one who believes it is imposing a burden.
The Commission has also been explicit that the rules were developed after wide consultation with businesses, NGOs and experts, with the stated aim of making them work in practice without creating unnecessary red tape. That claim will be tested by the first round of disclosures.
The Parliament: the model, not the waste
The parliamentary framing has always been more radical than the Commission's. Alessandra Moretti, the MEP who steered the file through Parliament, argued at the point of political agreement that the objective was to end the "take, make, dispose" model itself, positioning the destruction ban as one instrument among many rather than as a discrete waste measure.
This matters for forecasting. A Parliament that understands the ban as an attack on a business model will not be satisfied by compliance that keeps overproduction constant and redirects surplus into donation channels. It is the constituency that will press for scope extension at the 2028 review.
The critics: derogations as loopholes
The most substantive criticism of the measure concerns the derogations, and it predates their adoption. A written question tabled in the European Parliament in July 2025 raised precisely this concern: that broad exceptions, particularly those grounded in intellectual property or in claims that goods are unrepairable, could permit the continued destruction of functional products.
The Commission's own consultation record shows it took some of this seriously. Respondents warned that a cost-effectiveness criterion could create an incentive to destroy low-value items, that voluntary standards used as justification could open a route around the ban, and that donation protocols needed either tightening or loosening depending on who was asked. The Commission narrowed the voluntary-standards derogation to chemical safety and extended the donation window to eight weeks in response.
Whether that is sufficient is a live question, and it is the one to watch. If the first disclosures show heavy reliance on ground (f), damaged product, or ground (h), failed donation, the political pressure to tighten will be immediate. Both grounds are, not coincidentally, the two that a brand with weak evidence discipline is most likely to over-claim.
There is a second criticism, less often heard in Brussels and more often in the reuse sector: that donation itself has functioned as a disposal route, with large volumes of unsold European stock exported to markets in Africa and South Asia where it becomes somebody else's waste problem. The regulation does not directly address export of donated goods. Brands treating donation as the simple answer to the ban should expect that gap to attract attention within this legislative cycle.
The industry: volume of regulation, not the principle
The trade position has generally not been to defend destruction. It has been to warn about accumulation. Dirk Vantyghem, Director General of EURATEX, has for several years catalogued the number of sustainability files landing on the same companies at the same time, counting "Sixteen legislations on sustainability" in development affecting textile and clothing companies.
That is the honest industry concern, and it is a fair one. The destruction ban does not arrive alone. It arrives alongside mandatory textile extended producer responsibility under the revised Waste Framework Directive, the Digital Product Passport, the Green Claims regime, corporate sustainability reporting, and the deforestation regulation which reaches leather. Each has its own data model, its own retention period and its own competent authority. For a brand of £50 million turnover with a compliance function of one and a half people, the aggregate is the problem, not any single instrument.
The practical implication is that the brands who cope will be those who build one product data spine capable of serving multiple regimes, rather than one project per regulation. That is a technology architecture decision, and it is being made in most businesses this year by default rather than deliberately.
6. Implications for operations
This section names decisions and attaches dates. Financial year ends vary, so where a deadline depends on the reporting cycle it is expressed relative to your own year end.
Decision 1: confirm your classification, this month
The decision: whether you are a large enterprise for ESPR purposes as of 19 July 2026, assessed at group level including linked enterprises.
Why it cannot wait: the prohibition is already in force. If you are large and you dispose of Annex VII stock next week without a documented derogation, that is a breach on the day it happens, not a reporting problem next year.
What to do: have your finance director produce the headcount, expressed in annual work units, plus turnover and balance sheet position for the last two accounting periods, at group level. Aggregate holdings between 25% and 50% proportionally as partner enterprises, and holdings above 50% in full as linked enterprises. If the answer is large, this report is an operating manual. If the answer is medium, you have until 19 July 2030 for the ban, and the Article 23 prevention duty applies to you today regardless.
Deadline: within 30 days.
Decision 2: freeze routine disposal of Annex VII stock until the evidence process exists
The decision: whether to continue disposing of apparel, accessories and footwear on the existing process while the compliance process is built, or to hold stock.
The arithmetic: holding 4,000 units of dead stock for one quarter costs a fraction of a pallet-week rate. A breach of Article 25 costs an enforcement file, a potential market access intervention, and a disclosure that will be read against your competitors from 2027.
What to do: issue a written instruction that no Annex VII stock is disposed of without a named derogation and a completed evidence file, countersigned by whoever holds compliance. Where stock is held by a 3PL with standing disposal authority, withdraw that authority in writing this week. Standing instructions given to logistics partners two years ago are, in a number of businesses, the single largest live exposure.
Deadline: immediate, and in writing.
Decision 3: extend write-off reason codes from financial to legal, before your next season close
The decision: whether the stock adjustment reason code list is extended to carry the ten derogation grounds alongside the existing financial categories.
Why this specific change: it is the smallest possible intervention that produces a compliant record, and it lands in a system finance already controls. The alternative, reconstructing legal grounds retrospectively from value-based ledger entries at year end, is the failure mode this whole report is written to prevent.
What to do: map the ten grounds into the ERP reason code table. Retain the financial code as a second field rather than replacing it, because the audit and the disclosure need different things. Require both fields on any write-off of Annex VII product.
Deadline: before your next season close, and in any event before your financial year end, so the first affected reporting year is captured cleanly.
The reason-code mapping, including the validation rule that blocks the residual ground, is set out in full in The ten derogations: a working reference.
Decision 4: join the write-off record to the product master weight
The decision: whether to build the unit-to-weight join now or to attempt a weight reconstruction at reporting time.
The arithmetic: the weight already exists. PLM holds it from development, ERP holds it for customs and shipping. A unit-level write-off multiplied by product master weight produces the kilogram figure Annex I requires, at a tolerance well inside the 10% margin in Annex III. Reconstructing weight at year end from a value-based adjustment requires estimating an average garment weight across a mixed write-off, which will not survive the plausibility check.
What to do: specify the join as a reporting view, not a migration. Confirm which system is the master for weight and that the field is populated to acceptable completeness across live and discontinued styles, because discontinued styles are precisely the ones you will be writing off. Where completeness is below 90%, commission a backfill against the categories most likely to be discarded.
Deadline: specify in this quarter, live before your financial year end.
Decision 5: renegotiate the disposal clause in your 3PL and waste contracts
The decision: whether to amend third-party contracts to require treatment outcome reporting, or to accept a permanent unknown percentage in your disclosure.
Why this is urgent rather than important: Annex III names a high proportion of unknown treatment outcomes as a specific risk indicator that draws authority attention. A brand that cannot say what happened to its goods is flagging itself.
What to do: three clauses. First, no disposal of Annex VII product without written instruction referencing a derogation. Second, treatment outcome reported back per consignment, split across reuse, recycling, other recovery and disposal, referenced to your write-off number. Third, the waste operator statement required by Article 4 of the Delegated Regulation is issued and copied to you. Most 3PL contracts renew annually; get these into the next renewal rather than opening a variation.
Add a fourth point on retention, because it lands on the same contracts. The waste operator's statement and the delivery and reception documentation run on a five-year clock counted from the date of your disclosure, not from the date of collection, which is normally the later of the two. A contract that requires your partner to retain and reproduce these records for five years from collection will expire before your obligation does.
Deadline: at next contract renewal, or within 90 days, whichever is sooner.
Decision 6: start the donation route eight weeks before you need it
The decision: whether to build a standing donation channel now, or to seek one when stock needs to move.
The arithmetic that makes this non-optional: ground (h) requires either offers to at least three social economy entities in the EU or a public listing running a minimum of eight weeks. Eight weeks after the moment you decide to clear space is too late. It has to run in parallel with the final markdown phase, which means the decision to offer has to be made while the stock is still commercially live.
What to do: identify and qualify at least three social economy entities per major EU market now, with a written protocol covering what you will offer, in what condition, and who pays transport. Build the eight-week public listing mechanism as a standing page rather than a per-event exercise. Set the trigger point in the trading calendar: the offer starts at the same moment the final markdown is authorised, not after it fails.
Where the eight-week window sits in the sequence, and what happens when goods are damaged inside it, is worked through in One write-off, start to finish.
Deadline: partners qualified and protocol documented within 90 days. Trigger point built into the AW26 markdown calendar.
Decision 7: scope the disclosure project against the full product range, not just apparel
The decision: whether the Article 24 disclosure project covers only Annex VII goods or the full 50-plus category list in the Implementing Regulation.
Why brands get this wrong: the ban and the disclosure have different scopes. Homeware, bags and luggage, small electricals, licensed toys and tyres all appear in the disclosure categories. A brand with a home or accessories line that scopes its reporting to apparel and footwear will file an incomplete disclosure.
What to do: run your product master against the customs code list in the Implementing Regulation, at two-digit level, flagging any category requiring four-digit granularity under Annex II. This is a one-day exercise for anybody with access to your classification data and it prevents a structural error.
Deadline: before disclosure system specification, and in any event within the current quarter.
Decision 8: decide who owns this, and say so
The decision: which named individual is accountable for ESPR unsold goods compliance.
Why it matters more than it should: the obligation sits across finance, which makes the write-off decision, operations, which executes disposal, and compliance, which owns the evidence. In a business of £20 million to £100 million there is usually nobody whose job description covers all three, and the predictable result is that each assumes another has it.
What to do: name one owner, give them authority over the disposal instruction, and put the annual disclosure on the board calendar twelve months ahead of its due date rather than three.
Why the ownership question has no natural home in a group structure is taken up in Who is the economic operator?.
Deadline: named at the next senior team meeting.
The cost, honestly stated
None of the above is a large capital project. The five system changes described in section 4.4 are configuration and reporting work in systems you already own, plus contract variations. In a business with a functioning PLM and ERP, a competent internal systems lead and an engaged finance director, this is a matter of weeks of effort spread across a quarter, not a six-figure implementation.
The cost that is real, and that nobody should minimise, is the disposal cost that the ban transfers back onto the brand. Destruction was cheap because it was a single transaction. Resale, repair, donation logistics and preparation for reuse are each more expensive per unit than incineration, and the difference is not recoverable in a market where the goods failed to sell at any price. For a brand discarding, say, 30,000 units a year, the incremental handling cost is a genuine line item and should be budgeted rather than absorbed silently by operations.
That cost is also the entire point of the measure. It makes overproduction more expensive at the moment of disposal, in the hope that it becomes less attractive at the moment of buying. Whether that transmission actually works is the open question of the next five years.
7. Frequently asked questions
Does the ban apply to my business if I am based outside the EU? Yes, if you are a large enterprise and you place apparel, clothing accessories or footwear on the EU market. The obligations attach to market placement, not to establishment. UK, US, Swiss and Turkish brands selling into the Union are in scope on the same terms as EU-established ones.
We employ 180 people and turn over £45 million. Are we in scope now? Almost certainly not for the prohibition, which reaches medium-sized enterprises on 19 July 2030. Check group ownership before relying on that: if you are majority owned by a larger group, you are assessed with the group. And note the Article 23 prevention duty applies to you today regardless of size.
Is recycling still allowed? Only under a derogation. This is the most commonly misunderstood point in the regulation. Article 2(34) defines destruction as the intentional damaging or discarding of a product as waste, and consignment to a recycler is discarding as waste. Recital 3 of the Delegated Regulation lists what is excluded, and the list runs to donating, remanufacturing, refurbishing and preparing for reuse only. The Annex I disclosure template puts recycling inside the destruction group alongside other recovery and disposal. Sending unsold stock to a fibre recycler without a derogation is a breach. Where a derogation does apply, recycling must be prioritised over other recovery and over disposal.
What is not destruction? Donation, remanufacturing, refurbishment and preparation for reuse. These are the routes the regulation wants you to exhaust first, and using them does not trigger the derogation machinery at all.
How long must we keep the evidence? Two clocks, and they run from different events. Derogation evidence: five years from the destruction, in electronic form, produced to a competent authority within 30 days of request. Delivery and reception documentation, including the waste treatment operator's statement: five years from the date the disclosure was made, a rule stated consistently by Cooley, Bird & Bird and Beveridge & Diamond. The second clock is usually the later one.
We were briefed that documentation had to be kept for ten years. Which is right? Five. The ten-year figure was in the draft delegated regulation that went out for consultation in June 2025, and the final act halved it, as Linklaters records in its coverage of that consultation. Anyone working from 2025 consultation-era notes or an internal memo written at the time has the superseded number and will over-specify their archive.
Do we need a separate evidence file for every garment? No. The Delegated Regulation permits documentation to be prepared collectively where the same circumstances affect multiple products. Batch-level evidence is workable, provided the batch genuinely shares the same ground.
What counts as a valid donation attempt? Under ground (h), an offer to at least three social economy entities within the EU, or a public listing on your own website running for a minimum of eight weeks, in either case not accepted. Ground (h) is only available where grounds (a) to (g) do not apply, so you cannot use a failed donation to cover goods that were actually damaged.
What is a social economy entity? The Delegated Regulation defines it by reference to Article 3(4i) of the Waste Framework Directive. It is not a synonym for any charity. Qualify your partners against the definition before relying on them for compliance.
Our stock is damaged in the warehouse. Can we destroy it? Under ground (f), yes, if the damage makes it unacceptable for consumer use and repair is not technically feasible or not cost-effective, and you hold either documented quality assessment procedures or an inspection record. Note the definition of cost-effective: the comparison is against the cost of destruction plus full replacement, not against the item's residual sale value.
What about counterfeits and licensed product? Counterfeit and IP-infringing goods fall under ground (c), evidenced by a judicial decision, ADR outcome, rights holder notification or a substantiated internal investigation. Licensed product whose sale window has expired falls under ground (d), but you must also justify why destruction was proportionate rather than another route.
Do returns count as unsold stock? Yes, expressly. Article 2(37) includes products returned by a consumer under the Article 9 right of withdrawal, or during any longer withdrawal period the trader chooses to offer. Classification does not depend on whether the unit passes inspection or is restocked. The disclosure and derogation obligations bite only when the unit is actually discarded. Returns are the largest single source of exposure for e-commerce-led brands.
What exactly do we have to publish, and where? Under Annex I of the Implementing Regulation: entity identification, product data by customs code with units and weight, the waste treatment split across reuse, recycling, other recovery, disposal and unknown, and prevention measures both taken and planned. Published on your website, or in a CSRD sustainability report with a link from the website, within twelve months of your financial year end.
When does the standardised format start? Later than almost every summary says. The Implementing Regulation applies from 2 March 2027, and the prescribed format governs financial years starting on or after that date. A March year end means the first affected year runs to March 2028 and the disclosure is due by March 2029. The Commission's own press summary describes the format as applying from February 2027, without the financial-year qualifier, and the trade press has repeated it. What changes is only the format. The disclosure obligation itself has been running since the first full financial year commencing on or after 18 July 2024.
We have never published a disclosure. Are we late? Possibly, and this is the most urgent line in this report. Large enterprises have owed a public disclosure covering their first full financial year commencing on or after 18 July 2024, published within twelve months of that year end. For most brands that meant publishing during 2026 on FY25 data, in whatever format they chose. If you are a large enterprise and nothing has been published, the obligation has already been missed and the remedy is to publish, not to wait for the prescribed template in 2027.
When exactly does a return become an unsold consumer product? On return, not on inspection and not on a restock decision. Products returned by a consumer under the statutory right of withdrawal are expressly within the definition, and Article 9 of Directive 2011/83/EU gives 14 days from delivery on distance contracts. A brand offering a longer returns window has extended, by its own commercial choice, the period during which its returns carry this status. Classification does not depend on what your returns team does next; the obligations bite only when the goods are actually discarded.
The ESPR does not seem to define "large enterprise" anywhere. Have we missed it? No. The definitions article imports the definitions of SME, small and micro enterprise from Commission Recommendation 2003/361/EC and derives everything above them by subtraction. Searching the text for the phrase returns nothing, which has led several readers to conclude the test is missing or that another instrument must govern.
How accurate do our figures have to be? Annex III of the Implementing Regulation treats a discrepancy of under 10% between disclosed figures and supporting documentation as compliant. Above that, an authority may find non-compliance.
Which products are covered by the ban, precisely? Annex VII of the ESPR, identified by customs classification: apparel and clothing accessories in chapters 61 and 62, leather apparel and accessories under heading 4203, and footwear under headings 6401 to 6405. Verify your own range against the Annex text rather than against a summary, including this one.
We run 40 stores and employ over 250 people, but our turnover is only £35 million. Are we large? Possibly not, and it is worth checking properly rather than assuming. Under the Accounting Directive test you need to exceed two of the three thresholds, so headcount alone does not make you large if turnover and balance sheet are both below. Under the SME Recommendation, headcount is counted in annual work units, meaning part-time and seasonal staff count pro rata. A 300-name payroll across a store estate can convert to well under 250 annual work units. Run the calculation before you conclude anything.
We are online-only with 60 staff and a 3PL doing our fulfilment. Does outsourcing keep us out of scope? It may delay you, and that is an accident of the drafting rather than an intention. Because your warehouse and returns headcount sits on your logistics provider's payroll, your own employee count stays low, which under the SME Recommendation test would keep you medium-sized until 2030 regardless of turnover. Under the Accounting Directive test it would not help you, because turnover and balance sheet alone can put you over. Two things to check before relying on it: your group ownership, since a private equity or corporate parent is consolidated with you and will almost certainly make you large today, and the Article 23 prevention duty, which applies to you now whatever your size.
Is the disclosure scope the same as the ban scope? No, and this trips people up. The disclosure format covers more than 50 product categories including electronics, appliances, furniture, toys, tyres and home textiles. The ban currently covers only Annex VII apparel and footwear. Scope your reporting wider than your prohibition.
What are the penalties? Set nationally. Article 74 of the ESPR requires penalties that are effective, proportionate and dissuasive, enforced by member state authorities. The ESPR does not set an EU-wide 4% of turnover ceiling; that figure belongs to the EU Deforestation Regulation and has been misattributed in some commentary. The more serious commercial risk is market surveillance action restricting the making available of product.
How will authorities decide who to investigate? Risk-based, per Annex III. Trigger indicators include no disclosure at all, implausibly low figures for the size of the operator, a high proportion of unknown treatment outcomes, previous non-compliance, and cross-referencing against other data sources including tax declarations.
Do we have to tell our waste contractor anything? Yes. Article 4 of the Delegated Regulation requires you to provide the receiving waste treatment operator with a statement identifying the derogation relied on. This is a new obligation flowing outward to a supplier, and it is easy to miss.
Will the ban extend to other product categories? The ESPR empowers the Commission to add categories to Annex VII. The 2028 mid-term review of the working plan is the most likely moment. Brands with homeware, accessories or licensed hard goods should read the current measure as a preview rather than as somebody else's file.
How does this interact with textile EPR? The revised Waste Framework Directive, in force since 16 October 2025, requires member states to establish mandatory extended producer responsibility schemes for textiles and footwear, with fees eco-modulated against criteria developed under the ESPR. The destruction ban governs what you may not do with surplus; EPR governs what you pay for having placed the product on the market. They are separate obligations with overlapping data requirements, and they should be scoped as one data project.
Does the Digital Product Passport help with any of this? Not yet directly, since the DPP delegated acts for textiles are still ahead. But the data model overlaps substantially, and a brand building unit-level identification for DPP purposes is building the same spine that makes unsold goods reporting straightforward. Sequence the two together rather than separately.
We are a medium-sized brand. What should we actually do before 2030? Comply with Article 23 now, since the prevention duty has no size exemption. Then use the four years to do cheaply what large brands are doing under time pressure: unit-level write-offs, legal reason codes, a weight join, and contracts that report treatment outcomes back to you. The 2030 deadline is generous only if you use it.
8. Closing: the second answer
Return to the write-off code.
It was never a lie. It was an accurate description of a commercial reality, recorded in the only language the system that produced it could speak. Obsolete meant obsolete. The garment had no route to margin, and everything downstream of that judgement was logistics.
What the ESPR has done is insist that a second answer exists alongside the first. Not instead of it, because the commercial judgement remains valid and no regulation will make dead stock sell. Alongside it. The finance decision now needs an operational twin: not only what this stock is worth, but what physically happened to it, in units, in kilograms, on which legal ground, to which destination, evidenced for five years.
That is a small change and a large one at the same time. Small, because the data mostly exists already, scattered across a PLM that knows the weight, an ERP that knows the quantity, and a logistics partner who knows the destination. Large, because nothing in the way those systems were designed anticipated that anybody would ever need to join them, and because the join has to be made before the stock moves rather than after.
The brands that will find the next three years comfortable are the ones that treat the coming quarter as a records project rather than a sustainability project. The ones that will find it uncomfortable are those waiting to see how enforcement develops, on the reasonable assumption that a new regulation takes a year or two to bite.
That assumption is usually sound, and it is worth being precise about why it is unreliable here, because the sequence is counter-intuitive and almost nobody is discussing it. Standardised, brand-level, comparable disclosure arrives late. The prescribed format governs financial years starting on or after 2 March 2027, which for a March year end means the first disclosure in that format is published in March 2029. That is a longer runway than the trade coverage implies, and it should be used rather than feared.
But the Union-level picture arrives first. Under Article 25 the Commission must publish consolidated information on the destruction of unsold consumer products by 19 July 2027, and every 36 months after that. So the aggregate becomes visible roughly two years before individual brands are publishing in a common format. Anomalies at market level will be readable while individual disclosures are still in each operator's own presentation, which is precisely the condition under which a regulator goes looking for the operators whose numbers do not fit the total.
That makes the intervening period the cheap window rather than the safe one. The work described in this report costs weeks of configuration effort if it is started now and a great deal more if it is started when somebody asks for the file.
The interesting question is not whether brands will comply. Most will, because the cost of compliance is modest and the cost of exposure is not. The interesting question is what the first full year of comparable data reveals about the size of the problem, and whether a sector that has never had to publish this number discovers that it is smaller than the campaigners feared or larger than the boardrooms believed.
Both answers would change something. The second would change a great deal.
9. Companion articles
Three companion pieces expand specific sections of this report. Each is written to stand alone and each links back to the section it develops.
| Article | What it gives you | Expands |
|---|---|---|
| One write-off, start to finish: what a compliant record actually looks like | One consignment followed from returns bench to published disclosure line, with the cost-effectiveness arithmetic worked in full | 4.2, 4.4, Decision 5 |
| The ten derogations: a working reference for fashion operations and systems teams | Printable tables: the ten grounds with evidence and owner against each, a reason-code mapping to hand to IT, and the three obligations side by side | 4.2, 4.3, Decision 3 |
| Who is the economic operator? The scope questions the destruction ban leaves open | Which operator carries the obligation across wholesale, franchise, concession and third-party logistics models, and where the regime is silent | 4.1, 4.1a, 4.4 |
10. References and data sources
Primary legislation and official sources
1. Regulation (EU) 2024/1781 of the European Parliament and of the Council establishing a framework for the setting of ecodesign requirements for sustainable products (ESPR). Official Journal, 28 June 2024. https://eur-lex.europa.eu/eli/reg/2024/1781/oj
2. Commission Delegated Regulation (EU) 2026/296 setting out derogations from the prohibition of destruction of unsold consumer products. Adopted 9 February 2026, in effect 12 May 2026. https://eur-lex.europa.eu/eli/reg_del/2026/296/oj
3. Commission Implementing Regulation (EU) 2026/2 on the details and format for the disclosure of information on discarded unsold consumer products. Published 10 February 2026, in force 2 March 2026, applies from 2 March 2027. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ:L_202600002
4. European Commission, DG Environment. "Ban on destruction of unsold clothes and shoes enters into application." 17 July 2026. https://environment.ec.europa.eu/news/ban-destruction-unsold-clothes-and-shoes-enters-application-2026-07-17_en
5. European Commission, DG Environment. "New EU rules to stop the destruction of unsold clothes and shoes." 9 February 2026. https://environment.ec.europa.eu/news/new-eu-rules-stop-destruction-unsold-clothes-and-shoes-2026-02-09_en
6. European Environment Agency. "The destruction of returned and unsold textiles in Europe's circular economy." Briefing, published 4 March 2024. https://www.eea.europa.eu/en/analysis/publications/the-destruction-of-returned-and-unsold-textiles-in-europes-circular-economy
7. Directive 2013/34/EU (Accounting Directive), enterprise size classification. https://eur-lex.europa.eu/eli/dir/2013/34/oj
8. Directive 2008/98/EC (Waste Framework Directive), as revised, including the definition of social economy entity and the textile EPR obligation. https://eur-lex.europa.eu/eli/dir/2008/98/oj
9. Regulation (EU) 2023/988 (General Product Safety Regulation). https://eur-lex.europa.eu/eli/reg/2023/988/oj
10. European Parliament. "Exemptions to the ban on the destruction of unsold consumer products", question for written answer E-002831/2025, 14 July 2025. https://www.europarl.europa.eu/doceo/document/E-10-2025-002831_EN.html
11. Council of the European Union. "Ecodesign requirements for more sustainable products", policy page. https://www.consilium.europa.eu/en/policies/ecodesign-requirements-for-more-sustainable-products/
12. European Commission circular economy platform. "New EU rules to stop the destruction of unsold clothes and shoes." https://circulareconomy.europa.eu/platform/fr/node/9650
Legal and professional analysis
Cooley Productwise, "Deep Dive: EU Finalises New Requirements for Unsold Consumer Products Under ESPR", 7 May 2026. https://products.cooley.com/2026/05/07/deep-dive-eu-finalises-new-requirements-for-unsold-consumer-products-under-espr/
Bird & Bird, "EU: Destruction of unsold products, is your company Ecodesign compliant?", February 2026. https://www.twobirds.com/en/insights/2026/eu-destruction-of-unsold-products--is-your-company-ecodesign-compliant
Linklaters Sustainable Futures, "EU ESPR: Commission adopts final acts on unsold consumer product destruction and disclosure". https://sustainablefutures.linklaters.com/post/102miiy/eu-espr-commission-adopts-final-acts-on-unsold-consumer-product-destruction-and
Crowell & Moring, "Companies Selling Consumer Products in the EU (Particularly Clothing, Apparel, and Footwear Companies) Beware". https://www.crowell.com/en/insights/client-alerts/companies-selling-consumer-products-in-the-eu-particularly-clothing-apparel-and-footwear-companies-beware-the-new-espr-rules-on-unsold-consumer-products-have-now-entered-into-force
Commission Recommendation 2003/361/EC concerning the definition of micro, small and medium-sized enterprises. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A32003H0361
European Commission, ESPR Frequently Asked Questions, September 2024. Cited here only for its horizontal treatment of placing on the market. Its content on the unsold goods regime predates both secondary instruments and is not current. https://circabc.europa.eu/rest/download/25c48e7c-9ce3-41cb-96ac-d2942a8a29d6?ticket=
COSH!, "EU bans destroying unsold fashion: regulation or fast-fashion loophole?", March 2026. https://cosh.eco/en/articles/eu-legislation-espr-unsold-goods-ban
1. Generation Impact Global. "ESPR unsold consumer products: destruction ban and disclosure explained." 3 April 2026. Detailed treatment of the ten derogations, documentation requirements, the Annex I disclosure fields and the Annex III verification framework. https://generationimpact.global/blog/espr-unsold-consumer-products-destruction-ban-disclosure/
2. Generation Impact Global. "ESPR secondary legislation: delegated and implementing acts on unsold consumer products." https://generationimpact.global/blog/espr-secondary-legislation-delegated-implementing-acts-quick-reference/
3. Generation Impact Global. "EU cracks down on destruction of unsold fashion." Consultation record and stakeholder concerns. https://generationimpact.global/news/eu-ban-destruction-unsold-fashion-clothing-2025/
4. Baker McKenzie. "European Union: European Commission adopts new ESPR measures." 10 February 2026. https://www.bakermckenzie.com/en/insight/publications/2026/02/european-union-european-commission-adopts-new-espr-measures
5. Freshfields. "The 'E' of ESG: sustainable products, new EU ecodesign rules for unsold consumer products." https://www.freshfields.com/en/our-thinking/blogs/sustainability/the-e-of-esg-sustainable-products-new-eu-ecodesign-rules-for-unsold-consumer-102lqlo
6. Beveridge & Diamond. "EU clarifies ban on destruction of unsold apparel, clothing accessories, and footwear." https://www.bdlaw.com/publications/eu-clarifies-ban-on-destruction-of-unsold-apparel-clothing-accessories-and-footwear/
7. CMS. "New EU rules to prevent the destruction of unsold textiles." Includes the interaction with the revised Waste Framework Directive and textile EPR. https://cms.law/en/aut/legal-updates/new-eu-rules-to-prevent-the-destruction-of-unsold-textiles
8. Cattwyk Rechtsanwaltsgesellschaft. "ESPR: ban on destruction of unsold consumer products and disclosure requirements." Treatment of the Article 72(6) scrutiny period. https://www.cattwyk.com/en/news/espr-ban-on-destruction-of-unsold-consumer-products-and-disclosure-requirements
9. Anthesis. "ESPR: the ban on destruction of unsold goods." https://www.anthesisgroup.com/insights/espr-the-ban-on-destruction-of-unsold-goods/
10. Trustrace. "What happens on 19 July 2026 under ESPR." Treatment of recycling as destruction and the waste hierarchy within derogations. https://trustrace.com/knowledge-hub/what-happens-on-19-july-2026-under-espr
Sector and operational commentary
1. Ecosistant. "July 2026: EU ban on destruction of textiles and unsold goods." Accounting Directive thresholds and e-commerce returns implications. https://www.ecosistant.eu/en/ban-on-destruction-of-textiles
2. nShift. "EU ban on destroying unsold goods: what ESPR means for retailers." Reverse logistics and evidence-in-handoffs perspective. https://nshift.com/blog/eu-ban-on-destroying-unsold-goods-espr
3. Carbonfact. "ESPR crash course: how the Ecodesign for Sustainable Products Regulation will impact apparel and footwear brands." https://www.carbonfact.com/blog/policy/espr-textile
4. COSH! "EU bans destroying unsold fashion: regulation or fast-fashion loophole?" https://cosh.eco/en/articles/eu-legislation-espr-unsold-goods-ban
5. ESG Today. "EU bans destruction of unsold clothing, footwear." Commissioner Roswall's statement in full context. https://www.esgtoday.com/eu-bans-destruction-of-unsold-clothing-footwear/
6. WWD. "EU bans destruction of unsold clothes: new sustainable fashion rules." 10 February 2026. https://wwd.com/sustainability/business/eu-bans-destruction-of-unsold-clothes-sustainable-rules-1238556879/
7. Deutsche Welle. "EU ban on destroying unsold clothes takes effect." 19 July 2026. https://www.dw.com/en/eu-ban-on-destroying-unsold-clothes-takes-effect/a-78024888
8. World Footwear. "New EU regulations target destruction of unsold clothes and footwear." https://www.worldfootwear.com/news/new-eu-regulations-target-destruction-of-unsold-clothes-and-footwear/11252.html
9. Earth.org. "EU bans destruction of unsold clothes and shoes." EEA and Commission waste figures in context. https://earth.org/eu-bans-destruction-of-unsold-clothes-and-shoes/
10. Recover. "Bridging the gap: integrating recyclers' perspectives into European textile sustainability policies." EURATEX commentary on regulatory volume. https://recoverfiber.com/newsroom/bridging-the-gap-integrating-recyclers-perspectives-into-european-textile-sustainability-policies
Background: the French precedent
1. Coverage of the French Anti-Waste for a Circular Economy law (AGEC), adopted 10 February 2020, prohibiting destruction of unsold non-food products from 2022, and the MEP position on ending the take-make-dispose model. https://newsinfo.inquirer.net/1222955/france-cracks-down-on-plastic-waste-destruction-of-unsold-clothes
This report is published by fashion.business as an open resource. It is intelligence for operational planning, not legal advice. Verify your own scope and obligations against the regulation text and, where the stakes justify it, against qualified legal counsel in your primary market.
To continue reading, join fashion.business or sign in below.
Read the full report
Checking your membership…
