Textile EPR Fees: What Womenswear Brands Will Pay Per Garment From 2027

There is a particular kind of cost that arrives late and lands badly: a per-unit charge that was never in the costing sheet, applied to every piece you sell, set by somebody else, and influenced by decisions your designer made two seasons ago. Extended producer responsibility for textiles is that kind of cost.

It is also the most predictable surprise in the industry right now. The legal shape exists. The direction of travel exists. What does not exist yet is the number, and what will decide the number is partly in your hands — because eco-modulation means the fee is not flat. Two brands in the same country selling the same dress at the same price can pay different amounts because of how the garment was built.

That is why this belongs on a production desk rather than on a compliance desk. What follows is not a legal summary. It is the part of the mechanism that a brand can act on before the fee schedule is published: the design and construction decisions that move the band, and the data the producer will have to hold.

What EPR Actually Charges For

The clearest way to hold it is to say what it is not. EPR is not a waste disposal fee invoiced to you when your customer throws a dress away. It is a producer fee attached to the act of placing a product on a market. The trigger is the sale or supply into that country, not the end of life.

That single distinction changes the accounting. The cost accrues at the moment of sale, per unit, in the market where the sale happens. It therefore behaves like a duty or a levy in your landed cost model rather than like an after-sales provision — and it has to be forecast, not remembered.

It also means a brand selling into several Member States faces several regimes, with their own fee schedules, their own reporting formats and their own definition of who the producer is. For a brand that dropships or sells through a marketplace, the question of which entity counts as the producer in which country is the first one to settle, and it is not always the brand.

The Deadline, and Why It Is Not One Date

The revised waste framework rules entered into force in October 2025 and give Member States twenty months to transpose them into national law, which lands in the middle of 2027. From that point the obligation exists, but it exists once per transposing Member State rather than once for the bloc, each with its own schedule.

Two consequences follow, and brands routinely get both wrong. The first is that there is no single date on which textile EPR becomes payable across the EU. There is a transposition deadline, and then a set of national start dates, some of which will be earlier than others. A plan built around one date will be wrong in at least half the markets you sell into.

The second is that the fee schedule and the reporting format are national artefacts. What you have to declare, in what units, at what point in the year, and with what evidence, will differ by country. The data you need to produce, however, is largely the same everywhere — and that is the useful part, because it can be assembled once and reformatted many times.

So the practical posture is not to wait for the number. It is to build the file now, because the file is what every scheme will ask for, and the file is the slow part. A brand that has garment counts by material composition and by country can produce any format in a week. A brand that has to reconstruct that from invoices and packing lists will not.

Eco-Modulation: Where Construction Turns Into Money

Eco-modulation is the clause that turns this into a production subject. In its simplest form: the fee per unit is adjusted according to assessed environmental performance — durability, repairability, recyclability, and in some proposals the presence of substances of concern. A garment judged easier to keep in use and easier to recover pays less. One judged the opposite pays more.

It is easy to read that as a materials instruction — use better fibre, pay less. Fibre matters, but it is the smallest of the levers available to a designer. The larger ones are structural, and they are all decided before a sample is cut:

• Whether the garment is mono-material or mixed. A dress that is one fibre family throughout is simpler to assess and simpler to recover than the same dress with a mixed lining, a coated panel and elastane in the shell.
• Whether components can be separated. A zip, a boned bodice, a padded cup, a glued trim — each is a different material that has to be removed before the shell can be recovered. Removable beats bonded.
• Whether the garment is repairable. A hem that can be let down, a closure that can be replaced, a seam allowance that allows alteration: these are cheap to design and they are exactly what a durability assessment looks for.
• Whether finishes and coatings are declared. A performance finish changes how the material behaves at end of life and has to be disclosed, not omitted because the data is inconvenient.
• Whether the construction survives use. A garment that loses its shape in one season is doing the opposite of durability, regardless of what it is made of.

None of those five is a fibre choice. All five are decisions that a factory can implement and a brand can specify, and none of them requires waiting for the fee schedule to be published. That is the argument for acting now: the levers are known even though the rates are not.

The Design Choices That Move Your Band

Taken garment by garment, the same five levers show up as concrete specification lines. These are the ones worth reviewing on any style you intend to keep in range across the transition.

Two things about that table are worth saying out loud. The first is that almost every line is a decision you are already making — you are just not currently making it with this consequence in mind. The second is that none of these changes makes the garment worse. A removable cup is easier to service; a declared finish is easier to sell; an alterable seam allowance reduces returns. This is a rare case where the compliance-motivated choice and the quality-motivated choice point the same way.

Mono-Material Thinking Without Wrecking the Garment

The obvious objection is that occasionwear is the hardest category in which to ask for a mono-material garment. A structured evening dress exists because several materials are doing different jobs: a shell that drapes, a lining that moves, an interlining that holds an edge, bones that hold a bodice, a closure that takes tension. Asking for all of it to be one fibre is asking for a worse dress.

The workable version is not “one material”. It is separability plus disclosure. A garment built from four materials that can be taken apart and that are all declared is in a far better position than a garment built from two materials that are permanently fused and only one of which is documented. Assessment bodies know the difference between a composite and an assembly.

So the design question to put to a factory is not “can this be mono-material”. It is: what here is permanently bonded, and can it be captured instead; and what here is undeclared, and can we get the declaration. Those two questions will find almost all the avoidable exposure on a style, and neither requires the design to change.

What Data the Producer Has to Report

Every scheme needs a version of the same underlying facts. The format changes; the facts do not. Assemble these at style level and the reporting becomes a formatting exercise.

1. Units placed on the market, by country and by period. Not units shipped — units supplied into that market.
2. Material composition of each style, by percentage and by component. Shell, lining, interlining, trims, closures — each named, not a single headline fibre for the garment.
3. Mass per unit. Some schemes weight the fee by mass; a per-piece average is enough to forecast and is cheap to capture at packing.
4. Declared finishes, coatings and treatments, with the supplier declaration held against the style rather than in a mill's brochure.
5. Component and trim breakdown, particularly anything that is permanently bonded or that changes separability.
6. The evidence behind the durability and repairability claims you intend to make — test reports, construction notes, care instruction rationale.

That list will look familiar to anyone who has assembled a digital product passport file, and it should: the two regimes are asking for overlapping data about the same garments. Building them as one exercise rather than two is the single largest efficiency available here.

How It Lands on a Small Brand vs a Large One

The arithmetic favours the large brand and the pain falls on the small one, which is worth being honest about. A fee per unit is absorbed more easily at volume. A reporting obligation, though, is close to fixed cost: the same declaration is required whether you placed two hundred units or two hundred thousand.

The mitigation is to reduce the number of things you have to declare. A range of thirty styles each with a bespoke material stack is thirty files to assemble and maintain. A range built on three fabric platforms and four component sets is a much smaller set of declarations, most of which can be reused. That is a range-planning argument as much as a compliance one — and it is the same argument that makes a range cheaper to make.

There is also a threshold question in some schemes for very small producers. Where it exists it is defined nationally. It is worth checking early for each market you sell into, because it may determine whether you register at all — but it should be confirmed with the scheme or your adviser rather than assumed from another country's rules.

Where It Sits in Your Costing Sheet

A cost that arrives per unit, per market, at the moment of sale does not belong anywhere most costing sheets currently have a slot for it. It is not freight, because it is not about moving the goods. It is not duty, because it is not levied at import by customs. It is not a compliance overhead, because it scales with volume. It is closest to a duty in behaviour, and the safest treatment is to give it its own line rather than fold it into an existing one.

The reason to keep it separate is that a blended line cannot be re-forecast. If EPR is inside your landed freight percentage, then every change to a national fee schedule silently changes your freight assumption, and nobody will notice until the margin report looks wrong for reasons nobody can trace. A dedicated line, even with a placeholder value, means the assumption is visible, revisitable, and attributable when the number finally lands.

It also needs to be modelled per market rather than once. A brand selling into four Member States has four potential fee regimes, and averaging them produces a number that is wrong in all four. The unit economics of a style are already different by market once duty, freight and returns rate diverge; this is one more line in the same per-market view, not a new way of working.

There is a second-order effect worth putting in the model now, which is that eco-modulation may make two versions of the same style cost differently. If a construction change moves you into a better band, the saving is per unit and runs for as long as the style is sold. That is an argument for testing the change on one style before the schedules are published, because the comparison will be easier to read once there is a number attached to it.

What should not go in the model is a guess presented as a figure. A placeholder with a stated range and a stated source is honest and revisitable. A single invented number becomes the number everybody plans against, and it will be wrong.

The Same File Feeds the Digital Product Passport

The digital product passport and EPR are separate instruments with different purposes — one is about product data travelling with the item, the other about funding end-of-life management. But they are asking for the same underlying facts about the same garments, and the work of assembling those facts is the expensive part.

A brand that builds one structured product record per style — composition by component, declared finishes, mass, country of production, care rationale — has done most of both jobs. A brand that treats them as two projects will pay for the same data twice and end up with two inconsistent versions of it, which is worse than either alone.

The ordering is also convenient: the product data work is largely independent of the final schedules, so it can proceed now, while the fee modelling has to wait for national detail. Do the part that does not depend on the unknown.

What to Ask Your Manufacturer to Put in the File

This is the section a buyer can act on this week. None of it depends on a published fee schedule, and all of it is information a competent factory already has or can obtain.

Ask for it as a deliverable with the order, not as a favour at the end. A factory asked at quotation will price the documentation and collect it as it goes. A factory asked after shipment will reconstruct it from memory, and the reconstruction is what fails an audit.

Getting the Declaration Wrong

The failure modes here are dull and expensive, and they share a shape: the data exists somewhere in the business but not in the form the scheme wants, at the time it is wanted, attached to the thing it describes.

• A headline fibre instead of a component breakdown. The garment is declared as one material because that is what the fabric invoice says, and the lining, interlining and trims are never itemised. This is the single most common gap and it cannot be fixed retrospectively without opening the garment.
• A mill's generic certificate instead of a declaration for the batch. A certificate proves a facility once; a declaration describes what was actually used. Schemes and buyers ask for the second.
• Undeclared finishes and coatings. Omitted because nobody asked, or because disclosing them was assumed to create a problem. The omission is the problem.
• Packing lists used as a proxy for units placed on the market. Shipments are not sales, and a marketplace or distributor in between breaks the equivalence entirely.
• The file held by a person rather than by the style. The composition lives with whoever developed it, who has since moved on, and the style is still in range.

What follows from those five is usually not a fine on day one. It is a correction: re-declare, re-register, back-pay the difference, and in the worst case suspend supply into that market until the file is right. The correction costs more than the original disclosure would have, and it lands in the same season as everything else that went wrong.

The cheap defence is to attach the data to the style rather than to a person or a season. A per-style record that travels with the style through re-orders, recolours and re-runs is the thing that survives staff changes and scheme changes. It is also, not coincidentally, the same record that makes a re-order consistent with the original.

FAQ

How much will the fee actually be?

No honest answer exists yet, because each Member State sets its own fee schedule in national law and several have not published one. Anyone quoting you a per-garment figure before a scheme is published is guessing. What you can do now is reduce the number of construction choices that will push you into a worse band, and build the file.

When does it start?

The transposition deadline is in mid-2027, but the operative date is national, so there is no single start. Treat the middle of 2027 as the point by which your file should exist, not as the first day you pay. Confirm the actual start for each market with the scheme or your importer.

Is the brand the producer, or the importer?

It depends on how the supply chain is arranged and on national rules. Marketplace and distributor arrangements can move the obligation. This is worth settling in writing before you model the cost, because the answer determines who has to register and who forecasts the fee.

Does using recycled fibre reduce the fee?

Recycled content may be one of several modulated factors, but eco-modulation is broader than fibre — durability, repairability and separability are all in scope. A recycled shell bonded to an undeclared interlining is not a strong position. Treat fibre as one lever, not the lever.

Is this the same as the ban on destroying unsold stock?

No. That is a separate obligation under product ecodesign rules. They are adjacent in spirit — both push against disposal — but they have different triggers, different scopes and different evidence. Do not model them as one item.

We are outside the EU but sell into it. Does this reach us?

If you place products on an EU market, the obligation attaches to that act, not to your location. Expect to need a local representative or to work through whoever is treated as the producer in each market. Confirm the arrangement with your importer.

What is the one thing to do this quarter?

Build a per-style product record: composition by component, declared finishes, mass, and a note on what is permanently bonded. It is the same data every scheme will ask for, it overlaps with the digital product passport, and it is the part that does not depend on any number that has not been published yet.

Working With Luxudress

Luxudress is the factory-direct front end for womenswear production across production facilities in Guangzhou and Dongguan, covering development, sampling, material sourcing, manufacturing, inspection and export packing.

We will not quote you a fee, because no scheme has published one and we would be inventing it. What we will do is the part that sits with the factory: give you composition by component rather than a headline fibre, hold finish and coating declarations against your style instead of leaving them in a mill's file, tell you which parts of a garment are permanently bonded and which could be captured instead, and record packed mass per unit.

If you are planning a range that has to survive the transition, send the styles you intend to keep. You can also see how we work with rental businesses, read what the digital product passport will ask for, or see how we handle packing and labelling data.

Request a quote and we will return a costed plan including the material and construction documentation it assumes.