Recycled content
A certificate that arrives with a shipment usually records a volume relationship across a supply chain rather than the physical content of one garment. That is not a weaker version of the same claim. It is a different claim, and it decides what may be printed next to a product.
On this page
What a transaction certificate lets you say
A supplier sends a certificate. The question underneath it is always the same one: does this document say something about the units in this shipment, or does it say something about volumes across a period. Those are different documents doing different jobs, and only one of them supports a sentence about the garment a shopper is holding.
A mass balance record does not evidence physical recycled material in a specific unit. It records a relationship between inputs and outputs across a supply chain, so much recovered material in, so much certified output allocated out. The allocation is an accounting operation. It is not a trace, and a certified unit under that model may contain none of the recovered material at all.
Identity preservation is the other position. A physical lot is kept separate and identifiable from input through to product, and what comes out is the same material that went in. Where that holds, a statement about the unit is a statement the record actually supports.
Why the document is the only fact there is
Put a finished garment in front of a laboratory with a full instrument set. Composition comes back. Whether the polyester in it was recovered from a previous use or made from virgin feedstock does not, because no laboratory test at scale determines recycled from virgin origin. That proposition reaches us through a testing house summary of the preparatory study rather than from the study's own file, which is why it is recorded here as a second hand read.
Everything else on this page follows from that one sentence. Verification of a recycled claim rests entirely on chain of custody, because there is nothing else for it to rest on. The record is not evidence about the fact in the way a test report is evidence about a composition. For this field, the record is the fact.
That is more uncomfortable than the usual framing, and it should be. On most attributes the evidence has a hierarchy with a top rung: a test on the goods can contradict a certificate, and a certificate can contradict a declaration. Here the top rung is missing. Whatever the chain wrote down at the time is the whole of what is knowable afterwards, and a gap in that chain is not a weak claim but an absent one. Where each attribute becomes true, and how far upstream its evidence has to reach, is set out in how far back the evidence has to go.
Two different claims, not two grades of one
The most common error in this field is to treat mass balance as identity preservation with a weaker score, something to be qualified with a softer adverb and then published anyway. It is not that. The two records answer different questions, and no amount of caution turns an answer to one into an answer to the other.
| What you hold | What it establishes | What it does not establish |
|---|---|---|
| A mass balance record | A volume relationship across a supply chain, and that a certified quantity of output has been allocated against a certified quantity of input. | The physical content of any particular unit. A percentage taken from it and printed on a product page has changed subject without saying so. |
| An identity preserved record | That a named physical lot stayed separate and identifiable through the steps the record covers. | Anything about steps the record does not cover, or about units outside the lot it names. |
| A supplier declaration with no scheme behind it | What the supplier is willing to state in writing, with a date and an author. | That anybody outside the supplier has looked. It is evidence about what the supplier says, which is a real thing to hold and a smaller thing than it looks. |
Both terms above are used across the certification industry, and the definitions given here are ours, written so a reader can follow the argument. The scheme that issued your certificate defines its own terms, and its definition governs its own document.
There is a governed vocabulary for this, and we are not using it. An international standard on chain of custody names five models and defines each one, and those five governed model names sit behind a purchase this estate has not made. Every freely available account of them we found was a vendor page or a consultancy page. Paraphrasing a vendor's summary and presenting it as a standard's definition is precisely the failure this page exists to describe, so the two positions above carry our own definitions and our own label rather than borrowed ones.
What is free is the standard's published abstract, and it carries the sentence that matters more than the model names do.
Certificates in this field commonly come from voluntary schemes run by an industry body rather than from anything in law. Two of them, the Global Recycled Standard and the Recycled Claim Standard, are the ones a textile brand is most likely to be handed. We have not read either scheme text, so this page states none of their rules and describes none of their models. The right question to put to the certification body is narrow and they answer it well: which units does this document cover, and under which chain of custody model.
Where the arithmetic happens, and where it stops
Six schemes' published rules were read for this estate, and they carry more real control than a certificate on its own suggests. The central one is a volume reconciliation. Certified material in and certified output out have to be reconciled per lot or per batch, and a deviation trigger sits on top of that, which is a stated point past which a difference stops passing unremarked. What follows the trigger is each scheme's own procedure and this page does not restate it. Somebody is doing arithmetic, and the arithmetic has a threshold attached to it.
Then be exact about what the arithmetic establishes. It establishes that a certified operator's books balance across a lot, within a tolerance the scheme sets. It does not establish what is inside the units you received. Reconciliation is a statement about volumes, and the line on a product page is a statement about a garment, which is the same gap the first section describes arriving under a different name. The stated deviation figures live in each scheme's own document with its code, its version and its effective date, and they are not restated here from memory.
The scheme owners are also careful about what their own documents claim, and that care is worth borrowing. Two of the six rule sets we read print materially the same disclaimer on the face of a scope certificate: that it provides no proof that any goods delivered by its holder are certified. Proof of the goods is the transaction certificate. Two scheme owners, writing independently, chose to put that sentence on the credential itself.
We did not find, in the rules we read, a scheme owner stating that its certificate is disclosure for a product passport. If somebody tells you a certificate is passport ready, ask which scheme document says so and read that document. What each certificate does establish, and which of them carries a clock, is on which document is which.
The join between the evidence and the claim
Evidence arrives at the granularity the supply chain runs at. A mill certifies a lot. A recycler certifies a consignment. A converter certifies a run. Claims are made at the granularity a shop runs at, which is a product or a variant of one.
Those two granularities do not line up, and the mismatch is not occasional. One dye lot can be cut into two styles. One style can be produced in three runs from two mills across a season, with a certificate covering one of them. A colourway added mid season comes from whatever was available that week. Every one of those is ordinary, and every one of them breaks the join between the document and the product page.
Three
levels at which a passport may be established, model, batch or item
The framework leaves the choice to a product specific act and none has been adopted for textiles, so nothing yet fixes the level at which a textile claim would be made.
SourceEcodesign for Sustainable Products Regulation (EU) 2024/1781, consolidated textChecked 28 August 2026
Nothing we have looked at closes that gap for you. We went through the published passport data models asking a single question, whether any of them carries a rule for turning batch level evidence into a product level claim, and we did not find one. That is an absence found by us in the models we looked at rather than a statement that none exists, and the difference between those two sentences is set out on the evidence page.
Which level a passport would sit at is itself unsettled, and it is worth reading before this field is designed rather than after. The granularity guide covers what turns on model, batch and item. It also sets out what a business can settle for itself in the meantime.
What this field is allowed to say
Three states, and the passport says which one it is in. The middle row is where most of the work sits and it is the row that gets abused everywhere else.
| State | What it means | What produced it |
|---|---|---|
| Resolved | A proportion is stated for this unit, and the record behind it names the units it covers. | An identity preserved record, or a merchant's own production record joining a certified lot to the run this variant came from. |
| Partial | A claim exists about the input or about a volume, and it has an author and a date, and it does not reach this unit. | A mass balance record, or a supplier declaration. Attributed to whoever made it, and never promoted to the row above by rewording. |
| Not established | Nobody has stated a proportion for this product, or the document that exists cannot be joined to it. | The default. It is also where a mass balance record lands when there is no production record to join it to anything. |
Four things are never done to this field, and each of them is a thing done routinely elsewhere.
- No proportion is published that cannot be traced back to a document naming the units it covers. A number without that is a number about somebody else's material.
- The word verified is not used next to this field. It has no fixed meaning here, and it implies that a named party checked something against a named criterion, which is exactly the part that is usually missing.
- Recycled is not folded into the fibre name. It travels as its own claim, with its own evidence and its own state.
- Nothing is rounded up and nothing is renormalised. A proportion that arrived as a range stays a range.
A claim that reaches further than its document is not a strong claim with a small flaw. It is a different claim, with no evidence at all.
What a real platform asks for, and why it asks
Reading the law tells you what a claim must survive. Reading a sales channel's own field list tells you what a claim has to fit into, and the shape is more demanding than most internal records are. One platform's listing rules model recycled content as four things rather than as one number.
| What is asked for | What answering it forces you to hold |
|---|---|
| The attribute used | Which named standard or scheme the claim is being made under, rather than the bare word recycled attached to nothing. |
| The percentage of sustainable material, at product or at component level | A proportion with its level declared. The level is the part usually left implicit, and it is the part that decides whether the number is about the garment or about a piece of it. |
| Which components it applies to | A mapping from the claim to the parts of the product it covers, the shell or the lining or the trim, which is the same join the section above says nobody closes for you. |
| The certification level, its number and the testing institute | The identity of the document behind the number, recorded well enough that a third party could go and look at it. |
It is the most detailed recycled content field specification this estate has located anywhere, which is a statement about where we have looked rather than about the market. It is also carried at the weaker of two evidence tiers: it reached us through a sibling research stream rather than from the platform's own published policy, it has not been re-verified at source, and the platform is not named here for that reason. Treat it as the shape of a serious specification, not as a current rule you can be held to.
Now the part that sets expectations properly. On the platform material read for this estate, sustainability data behaves as a badge. It opens placement and promotion, and the platform's own rules describe the standards behind it as voluntary. Nothing read here makes this data a condition of supply or of listing, and a brand that fills the fields in is buying merchandising rather than permission to trade.
That is worth knowing before a programme is scoped around it. A badge is a commercial decision that can be taken slowly and reversed, and it does not carry the consequence a legal duty carries. It also has one useful property: filling in the fields honestly exposes exactly which joins in your own records are missing, which is work you would have to do anyway.
What this means for a catalogue
The useful work here is joining, not collecting. Most brands already hold more certificates than they can attach to anything, and adding another one does not move the field out of the middle row.
Start by asking each certificate what it covers, using those words, then record the answer next to the file rather than in somebody's memory. Then record which production run each variant came from. Where those two records meet, a claim about the unit becomes possible. Where they do not, the honest output is a claim about the input, stated as one.
Then read the words you actually publish against the scope of what you hold. A statement about the whole product resting on a fact about part of it is one of the practices the amending consumer information directive puts on the list of practices treated as unfair in every case, and a recycled content line is one of the easiest places in a catalogue to write one by accident. Whether the evidence in your folder supports the words on your product page is the question worked through on claims and substantiation.
The thing not to do is to publish the number on the certificate because a number looks better than a range and much better than an empty box. The certificate is somebody else's statement about somebody else's units, and once it is under your brand on a product page it has become yours.
Check what your own claims policy cites
A standard can be withdrawn without anybody telling the people who cite it. The international standard on self declared environmental claims had its edition withdrawn on 24 June 2026 and replaced by a new one. A claims policy, a supplier contract or a technical pack naming that edition now cites an edition that is no longer current, and nothing in the ordinary running of a business would surface it. The replacing edition sits behind a purchase, was not bought and was not read here, so nothing on this page describes what it contains.
Where the evidence has to come from
Nothing about recycled content can be read off the product, inferred from the fibre or reconstructed later. It exists only if somebody upstream wrote it down at the time, which makes this one of the fields where asking early is the whole method. How far up the chain that reaches, and where it is reasonable to stop asking, is on how far back the evidence has to go.
You might want to read next
Sources
-
Art. 5(1), Annex ICELEX 02011R1007-20180215In forceRelevant provisions reviewed
Read for the closed list of permitted fibre names and the rule that a name outside that list may not be used as a fibre name. It is on this page for a narrow reason: nothing we read in it addresses where a fibre came from, which is why a recycled claim is a second claim rather than part of a composition.
-
Art. 9(2)(d)CELEX 02024R1781-20240628In forceRelevant provisions reviewed
The framework leaves it to a product specific act to say whether a passport is established at model, batch or item level. That single provision is the reason the granularity question on this page is open rather than answered, and no such act has been adopted for textiles.
-
ProposedReached through a secondary reproduction, primary text not read
The preparatory study informing the textile decision, dated 13 May 2026. Its granularity section was read at its own address in an earlier pass. The proposition this page takes from it, that no laboratory test at scale determines recycled from virgin origin, reaches us through a testing house summary rather than from the study's own file, so the read is recorded as second hand and the section that uses it says so in the sentence that carries it. The study is pre-decisional by design and it is never cited here as law.
-
Chain of custody scheme rules, scope and transaction certificate specificationsScheme rulesRelevant provisions reviewed
Published scheme rules read at the schemes' own documentation. Two propositions on this page rest on them: that a volume reconciliation is required per lot or per batch with a deviation trigger attached to it, and that two of the six scheme rule sets read print materially the same disclaimer on the face of a scope certificate. The stated deviation figures sit in each scheme's own document with its code, its version and its effective date, those identifiers are held in the research record awaiting transcription, and no figure is written here from memory. Some of these documents carry effective dates later than the date at the top of this page, so the rows resting on them are written to be re-read rather than assumed.
-
ISO 22095, chain of custody, published abstractPublished standardPublished abstract read, full text behind a purchase
One sentence is taken from the published abstract and no more: a chain of custody document is not, on its own, able to support claims about an organisation's materials or products. The five governed model names sit behind a purchase this build did not make, every freely available account of those models we found was a vendor page or a consultancy page, and none of them is named here. It is the reason the two positions this page describes carry the page's own definitions rather than the standard's.
-
AdoptedOfficial source confirmed, detailed review pending
The amending directive on consumer information. It is named here for one enumerated practice, stating something about a whole product on the strength of a fact about part of it. The enacting terms were reached through this estate's research record rather than at the Official Journal, so the effect is described and no wording is quoted.
-
ISO 14021, self-declared environmental claims, edition withdrawn 24 June 2026Withdrawn 24 June 2026Official source confirmed, detailed review pending
Used for one fact, and it is a fact about the catalogue rather than about any text: the edition on self declared environmental claims is recorded as withdrawn on 24 June 2026 and replaced. A claims policy or a technical pack citing the withdrawn edition now cites an edition that is no longer current. The replacing edition is behind a purchase, has not been bought and has not been read, so nothing here describes what it contains.
-
Marketplace and platform seller policy documentationPlatform policy, not lawReached through a secondary reproduction, primary text not read
Behind the platform section only, and carried at the weaker of the two evidence tiers this material supports. The recycled content field shape and the finding that sustainability data behaves as a placement badge rather than as a condition of supply reached this estate through a sibling research stream rather than from a platform's own policy page, and neither was re-verified at source. The platform is therefore described rather than named, no wording is quoted from it and nothing here supports a statement that any marketplace requires passport data.
-
Industry modelRelevant provisions reviewed
One of the published passport data models this estate has looked at while asking whether any schema carries a rule for aggregating batch level evidence into a product level claim. The sentence on this page records what we did not find in the models we looked at, which is not the same as a statement about what any one model contains.
-
Voluntary schemeOfficial source confirmed, detailed review pending
Named so that a reader holding a certificate knows whose document it is and who to ask about it. Neither scheme text was read in this pass. No threshold, no chain of custody model and no rule of either scheme is stated anywhere on this page, and the scheme owner is the authority on what its own certificate covers.
Help someone else make sense of product passports.