Your chemical data describes a factory. Your obligation describes a garment.

The data is real. It is about a different object.

If your business has spent years on chemical management, the data you now hold is real, well governed and almost entirely about an organisation, a site, a process or a mixture. The passport question is about one garment, and those are different objects.

That is not a criticism of the programme. It was built to keep restricted substances out of production and to keep a site inside a standard, and it does that work. It was not built to carry a fact down to one item on a shelf, because until recently nobody was asking it to.

So the difficulty here is a join rather than a quality. A statement about a facility is not by itself a statement about an item that left it. A statement about a formulation is not a statement about the article the formulation ended up in. The programme and the passport field are both sound and they are pointed at different things.

The useful question is therefore not whether your scheme is a good one. It is which object your scheme's document attaches to, and whether anything joins that object to the garment you are about to publish a value against.

Six objects, and only one of them is the garment

Chemical data in textiles attaches to one of six objects: a facility, a formulation, a process, a material, a transaction or a product. Almost every document in the layer names one of the first five.

Drawing it is faster than describing it, because the picture shows why a complete programme can sit alongside an empty passport field without either of them being wrong.

The broken connectors are doing careful work. Each one marks a join we have not established rather than a join we have shown to be missing, and the difference between those two states, along with the rest of the vocabulary this estate uses for absence, is set out in how we know.

What each class of document certifies, and about what

The final column is a read status and it applies to the scheme documentation behind the row, not to the certificate in your folder. Where it says not read at source, the middle column describes what practitioners use the document for rather than what the scheme's own rules say, and that row makes no statement at all about what the scheme does or does not publish.

Schemes are named here by what they certify rather than by brand, for the same reason. Naming a scheme is a claim about that scheme, and a claim about a scheme needs its documentation read.

Two of those six rows are partly read and four are not read at source. Partly read means we have read some of a scheme's published rules at the scheme's own documentation and not all of them, so the row can describe the class and cannot describe its limits. Not read at source means we have read none of them, and the row is a description of usage.

The table compares objects and not quality. A document that attaches to a process is not a weaker document than one that attaches to an article. It is a document about a process, it is the right instrument for the question it was written to answer, and the certification bodies that maintain these schemes are the authority on what their own rules say. This page is about which object you are holding, which is a question you can settle in an afternoon and which almost nobody asks.

What this page is waiting on, and what would improve it

This page is gated on a read that is not finished, and it says so rather than writing around it. The proposition underneath it is a claim about what a set of documents does not do, and a claim of that shape is worth exactly what the read behind it is worth.

What remains outstanding is specific. It is the published rules of each class in the table above, at each scheme's own documentation, with the document code, the version and the effective date of the rules the answer rests on. Inside those rules there is one question that matters more than the rest: whether any scheme in this layer defines a route from its certificate to an individual article. Until that is read, no row here says what a scheme fails to do, and the connectors on the right of the diagram stay broken.

There is a better outcome available and it is worth naming plainly. If the completed read shows that a scheme does make a product level join, that is a better answer than this page, and we will publish it as one rather than defending the version you are reading. The finding would be more useful to you than the caution, and the page would be rewritten around it.

Where the join breaks, and where it does not

A restricted substance list is an instruction. It tells a supplier what not to use and at what limit, and holding one is good evidence that you gave the instruction. It is not a record of what a finished garment contains, because an instruction and an outcome are different facts and the gap between them is where the whole subject lives.

A wet processing conformance is a statement about a site across a period. If your garment was dyed at that site inside that period, the conformance describes conditions your garment passed through, which is genuinely useful and is not nothing. It is still a statement whose subject is the site. Turning it into a sentence about your garment is the specific error this page exists to name.

A tested article certification is the closest thing in this layer to a product record, and it carries its own scope even so: the article that was submitted, the panel that was applied and the terms the certification sets for itself. Whether it establishes anything about the specific item you are about to publish depends on what its own rules say about that, and those rules are on the outstanding list in the section above.

Nine parts out of ten is not a garment

Composition defeats the arithmetic even where every part is documented. A claim covering nine of ten components is not a claim about the assembly of ten, and we have not found a standard that defines how component level claims compose into a claim about a finished article. The tenth part is usually a thread, a coating, a trim or an adhesive. The passport field is about the object as sold.

That is why a folder that looks complete can still support nothing at the level a passport publishes at. What each kind of document does establish, and what ends its useful life, is set out on certificates, declarations and test reports.

There is one more join worth knowing about before you go looking for it. A reference to the public substances database can be recorded by whoever made the notification and cannot be discovered by anybody starting from a product, for reasons that are properties of the database rather than of anyone's search. That argument, and the five properties behind it, are on the SCIP reference page.

The one route that creates evidence

One route in law produces information that does not otherwise exist, and it is narrower than the problem it is usually reached for. The chemicals regulation's supplier information article has two paragraphs, and they are not the same duty pointing in two directions.

The cap at the end of that is the sentence that matters here. The duty compels disclosure of what a supplier already holds. It does not compel anybody to generate a fact about your garment that nobody has established, which is why a chemical programme cannot be converted into product evidence by asking harder or by asking more often.

How that route works in practice, what an unanswered request is worth and what to record when nobody replies are set out on getting evidence out of suppliers.

The clock underneath all of it

The list of substances the question is asked against moves, and nothing in a folder moves with it.

Every answer of the form no listed substance above the threshold is an answer about the list as it stood on the day it was given. It does not announce that it has aged. A statement that was correct at one revision may be describing a different list two revisions later, and the document carrying it will look exactly the same.

So the revision matters as much as the answer. Record which revision of the list a statement was made against and who made it, and treat an undated substances answer as an answer you cannot place in time. That is cheap to do at the point the document arrives and expensive to reconstruct afterwards.

What to ask your scheme

These are questions for the scheme rather than questions about it. They are the ones whose answers settle whether a certification attaches to the object you need, and they are answerable from documentation the scheme already publishes.

1. Which object does this certificate attach to? An organisation, a site, a process, a formulation, a material, a shipment or a finished article. Ask for the answer in the scheme's own words rather than accepting an inference from the certificate's title.
2. Do your rules define a route from this certificate to an individual article? If they do, ask which clause and read it. This is the question that changes the answer for everything else on the list.
3. Ask for the document code, the version and the effective date of the rules that answer rests on. Rules change, and an answer without its version cannot be checked later or defended later.
4. Ask what the scheme itself says this certificate does not cover. Ask for that statement in the scheme's own document rather than in a summary. It is usually shorter and clearer than anything a reader would infer.
5. What ends this certificate, and is anything downstream told when it does? A validity period a calendar can watch and a dependency nobody announces are different problems.
6. Ask which revision of the substance list the underlying assessment was made against. If the assessment does not name a revision, that is itself the answer.
7. If the certificate covers components, what do your rules say about an article assembled from certified and uncertified parts? Most disappointments in this area are hiding in that gap.

The first two settle most of it. If the scheme says the certificate attaches to a site, you are holding a site fact, and the right thing to do is to record it accurately as a site fact rather than asking it to be something else. If the scheme says it attaches to an article and points you at the clause, you are holding a great deal more than you thought, and that clause is worth reading twice.

Either way, the object your evidence attaches to belongs on the record next to the evidence itself. What the passport asks for, attribute by attribute, is on the attributes hub.

Sources

• Published scheme documentation for the textile chemical management data layer
  Scheme documentationOfficial source confirmed, detailed review pending

  The spine of this page and its publication gate. Partly read: seventeen of the twenty two observed sources behind the proposition have not been read, so every row of the scheme table carries its own read status and no row states what a scheme fails to publish. Where a row is unread it describes what practitioners use the document for rather than what the scheme's own rules say.

• REACH Regulation (EC) No 1907/2006, Articles 3(33) and 33
  Art. 33(1), Art. 33(2)CELEX 02006R1907-20250422In forceRelevant provisions reviewed

  Behind the route section only. Both paragraphs were read at the consolidated text, which is what allows this page to state that one limb pushes on supply, that the other is a consumer right with a stated period and that both are capped by the information available to the supplier. There is no statutory route by which a brand demands information from its own supplier, and the page is written to that correction.

  View official source

• ECHA Candidate List of substances of very high concern
  34th update, 253 substancesRegulator dataReviewed in fullChecked 21 August 2026

  Behind the clock section. Retrieved through the agency's own interface and revised roughly twice a year, which is why an answer about listed substances is treated here as an answer about a particular revision rather than a standing fact.

  View official source

• ECHA SCIP database public interface
  Regulator dataReviewed in full

  Carried for a two sentence pointer and nothing more. The interface properties that make a reference recordable and not discoverable were established against the public database itself and the argument belongs to the field page, which this page links to rather than restates.

  View official source

• Ecodesign for Sustainable Products Regulation (EU) 2024/1781, consolidated text
  CELEX 02024R1781-20240628In forceRelevant provisions reviewed

  Behind one sentence: the framework creates the passport and leaves the contents of a given product's passport to acts made under it. Articles nine to fifteen were not read verbatim for this page, so nothing here states what a textile act will contain or when anything applies.

  View official source