UKAS Accreditation: What the Mark Proves, and What an Unmarked Claim Does Not

A certificate of analysis arrives with the material. Somewhere on it, usually in small type near the footer, is a line about the testing laboratory. Sometimes that line says the laboratory is accredited and gives a number. Sometimes it says the laboratory works “in accordance with ISO 17025”, or “to ISO 17025 standards”, and gives nothing else.

Those two statements look almost identical on the page and mean very different things. The first is a claim somebody else has audited. The second is a claim about itself.

This is not a pedantic distinction. It is the single fastest check available to a buyer who cannot visit the laboratory, cannot repeat the test, and has to decide whether a number on a page is evidence or decoration.

What accreditation is

ISO/IEC 17025 is the international standard setting out general requirements for the competence of testing and calibration laboratories ^[2]. It covers the things that determine whether a result means anything: the competence of the people, the suitability and calibration of the equipment, the validation of the methods, the handling of samples, the estimation of measurement uncertainty, and the way results are reported.

The standard is a document. Anybody can buy it, read it, and organise their laboratory around it. Nothing about doing so involves an outside party, and nothing about it produces a mark.

Accreditation is the separate act of an independent body assessing a laboratory against that standard and formally recognising that it meets it. In the United Kingdom that body is UKAS, the national accreditation body, which assesses laboratories against ISO/IEC 17025 and publishes what it has accredited them to do ^[1].

The difference matters because assessment is where claims meet evidence. An assessor visits, examines records, watches methods being run, checks calibration chains, and reviews how the laboratory handles results it does not like. A laboratory that has been through that and a laboratory that has read the standard can both describe themselves as working to ISO/IEC 17025. Only one of them has been checked.

Accreditation is bounded by scope

The most commonly missed feature of accreditation is that it is never general. A laboratory is not simply “accredited”. It is accredited for specified activities, using specified methods, on specified materials, within specified ranges. That set of permissions is its schedule of accreditation, and UKAS publishes it ^[1].

This has an immediate practical consequence. A laboratory may hold accreditation for a technique it uses constantly and none at all for a technique it uses occasionally. A report from that laboratory can therefore contain both accredited and non-accredited results, and a well-run laboratory will mark which is which — typically by flagging the non-accredited lines or by stating that certain determinations fall outside the scope.

So “the laboratory is UKAS accredited” is not the end of the check. The question is whether it is accredited for the determination on your certificate. A laboratory accredited for one analytical technique has demonstrated competence in that technique; it has not thereby demonstrated competence in every other technique it happens to own an instrument for.

Why an unaccredited result is not automatically a bad one

It is worth being fair about this, because the opposite error is also common. An unaccredited result is not necessarily wrong, and treating accreditation as a binary of trustworthy and untrustworthy will lead you to reject good data and accept bad.

Plenty of competent laboratories run methods outside their accredited scope, for perfectly ordinary reasons: the method is new, the volume does not justify the assessment cost, or the determination is genuinely uncommon. Accreditation is expensive and slow, and a laboratory chooses where to spend it.

What accreditation changes is who is standing behind the claim. With it, an independent body has examined the method and the records and put its own name to the judgement. Without it, you are relying on the laboratory’s own account of its competence — which may be entirely accurate, and which you have no independent way to test.

For a buyer, that is the whole point. You are not in a position to assess a laboratory. Accreditation is the mechanism by which somebody who is in that position has done it for you, and published the result ^[3].

Verifying a claim in about two minutes

The check is deliberately easy, because accreditation is worthless if nobody looks it up. Every national accreditation body maintains a public register precisely so that a claim on a report can be tested by the person holding the report.

1. Find the accreditation number on the certificate. It is usually near the laboratory’s name or in the footer, often alongside a mark.
2. Look the number up on the accreditation body’s public register. For UK laboratories that is UKAS.
3. Check the organisation named on the register is the organisation named on your certificate — not a parent company, not a similarly named subsidiary.
4. Open the schedule and find the determination you care about. This is the step people skip, and it is the one that matters.
5. Check the accreditation is current. Accreditation can be suspended, withdrawn, or reduced in scope.

If the number does not resolve, or resolves to an organisation with a different name, or resolves to a schedule that does not include the determination on your certificate, you have learned something specific and useful. None of those findings tells you the result is wrong. All of them tell you the report does not carry the weight it appears to.

What accreditation does not tell you

Accreditation speaks to the competence of the testing. It says nothing whatever about the material.

A UKAS-accredited laboratory can produce an impeccable report on a sample that has no traceable relationship to the batch you received. It can test a sample that was selected, prepared and submitted by the seller. It can report exactly what it measured while being entirely silent about what it was not asked to measure.

Those are separate questions, and they are the ones a buyer usually needs answered next: whether the certificate is bound to your lot, whether the tests are dated before shipment, and what the analysis did not cover. Accreditation is a check on the laboratory. It is not a check on the chain of custody, and it is not a substitute for reading the rest of the document.

Treated as one check among several, it is a good one — fast, independent, and difficult to fake, because the register is public and the claim is falsifiable in a way that most supplier assurances are not.