Last verified 19 August 2026.
Why the number on the certificate is a different kind of number
Everything else you will read about a material is a statement about a population: a standard’s specified minimum is a floor the population must clear, a datasheet typical value is what the population usually does.
A 3.1 certificate is a statement about your material. It carries a heat number, and that heat number ties the chemistry and the mechanical results on the page to the plate on your floor. On the evidence ladder it is the strongest thing you will ever hold, and it is the only value that can retire a specified minimum rather than merely restating it.
Which is why it matters that three of the four documents in EN 10204 are not that.
The four documents
2.1 — Declaration of compliance with the order
The manufacturer states that the product meets the order requirements. No test results.
It tells you somebody at the mill is prepared to say the material conforms. It does not tell you what any test measured, because it does not contain any. Treat it as a commercial document, not as data.
2.2 — Test report
Actual test results, from non-specific inspection. Meaning: the tests were run on material from the production process, but not necessarily on the pieces you received.
This is the type most often mistaken for a 3.1, and the mistake is understandable — it has numbers on it, and the numbers are real. What it lacks is the link between those numbers and your delivery. A 2.2 tells you what the mill’s production is like. It does not tell you what your plate is like.
3.1 — Inspection certificate, specific inspection
Test results from inspection on the delivered material, validated by the manufacturer’s authorised representative independent of the manufacturing department — in practice quality assurance or a test house supervisor, not the production line.
A 3.1 typically carries: manufacturer name and address, the heat or cast number, the product description, chemical composition, mechanical test results, the order reference, and the authorised representative’s name and signature.
This is the document to ask for by default. It is the industry norm for anything structural or pressure-bearing, and the incremental cost over a 2.2 is usually small compared with the cost of not being able to answer a question about provenance later.
3.2 — Inspection certificate, countersigned
A 3.1 plus verification by a party independent of the manufacturer: your own representative, or an inspector designated under official regulations or by an accredited body.
Rare in general commerce, because it costs more and takes longer. Standard in some regulated sectors and in some purchase specifications. If a 3.2 is required, it will be required by the code or by the contract, and it will be explicit.
The four in one table
| Type | Test data | Inspection | Validated by | Ties to your material |
|---|---|---|---|---|
| 2.1 | none | — | manufacturer | no |
| 2.2 | yes | non-specific | manufacturer | no |
| 3.1 | yes | specific | manufacturer’s representative, independent of production | yes |
| 3.2 | yes | specific | as 3.1, plus an independent third party | yes |
Current edition: EN 10204:2004, superseding the 1991 edition. Definitions above are summarised from secondary sources; the standard itself was not consulted for this page, so under the TRACE demotion rule this is E3 content. Confirm against the standard before it governs a purchase specification.
What to actually check when one arrives
Six things, roughly in the order that catches problems fastest.
- The type. Printed on the document. If it is not printed, ask — a certificate that will not say what type it is has told you something.
- The heat number, and whether it matches. The number on the certificate should appear on the material, on the packing list and on your goods-in record. A 3.1 whose heat number matches nothing is a 2.2 with better formatting.
- The product form and dimensions. Certificates get reused across a mill order. Check the thickness and form on the page against what you received, because properties are specified by product form.
- The specification invoked. Which standard, which grade, which edition. This is where the two-standard problem shows up: a plate certified to EN 10088-2 and a plate certified to ASTM A240 clear different floors.
- The condition. Annealed, solution treated, quenched and tempered, cold worked. Mechanical results belong to a condition and mean nothing without it.
- The signature. A 3.1 needs a named authorised representative. A blank or an illegible stamp is worth a phone call.
What a certificate does not prove
It does not prove the material is fit for your application. It proves the material is what the specification says it is. Fitness is a design decision and stays yours.
It does not cover what was not tested. A 3.1 reports the tests the specification requires. It is silent on toughness at your service temperature, on corrosion in your medium, and on fatigue, unless those were ordered.
It does not survive processing. Values on the certificate belong to the material as delivered. Welding, forming, machining and heat treatment change the material, and the certificate does not follow it through your shop.
It is not automatically genuine. Certificate falsification exists and periodically surfaces in the trade press. Heat-number cross-checking against the mill, and buying through supply chains you can trace, are the practical defences.
Where it lands on the evidence ladder
| Document | Level | Why |
|---|---|---|
| 3.1 or 3.2 for your heat | E5 | Certified test result on the material you hold |
| 2.2 test report | E3 | Real data, no link to your delivery, method context thin |
| 2.1 declaration | E0 for data purposes | Contains no test results to grade |
Once a 3.1 arrives, any datasheet value you were carrying for that property is superseded and the record should say so. That swap is cheap to write down at the time and expensive to reconstruct later — see documenting material decisions.
What this page does not cover
- Certificates outside EN 10204. ASTM, ASME and JIS have their own documentation frameworks; the underlying distinction between specific and non-specific inspection recurs, the names do not.
- How to audit a mill. Different discipline.
- Detecting falsification beyond basic heat-number cross-checking.
- Non-metals. Polymer, composite and elastomer supply documentation is structured differently.
- The exact clause wording. This page summarises; the standard governs.
Sources consulted, both secondary: testcert.co EN 10204 reference · Holland Applied Technologies, 3.1 vs 3.2