What an accredited laboratory means: ISO/IEC 17025, ESYD and ILAC

"Accredited laboratory" is not a quality title and it is not a synonym for "certified". It means that a national accreditation body — in Greece, ESYD — has assessed the laboratory on site for specific measurements, using named methods, on specific materials, and has attested that competence in writing against ELOT EN ISO/IEC 17025, the Greek adoption of the international standard ^[1]. The attestation does not cover the laboratory as an entity; it covers a list. So when a test report is in front of you, the useful question is not "is the laboratory accredited?" but "was this method, on this material, inside the scope of accreditation on the day the report was issued?"

The distinction is not verbal. ISO/IEC 17011, the standard the accreditation bodies themselves operate under, defines accreditation as third-party attestation related to a conformity assessment body, conveying formal demonstration of its competence, impartiality and consistent operation in delivering specific activities ^[2]. Three words there do all the work: competence, impartiality, specific. A management-system certificate attests none of them; it attests that a documented system exists and is followed, not that the number coming off a chromatograph is right.

Accreditation, certification, declaration: three words that are not synonyms

In commercial copy the three terms are used interchangeably, and the confusion is not innocent: each says materially less than the one before it while sounding equally heavy. Accreditation assesses technical competence and is granted by a body that is itself subject to peer evaluation. System certification assesses procedures and is granted by a private body operating in a competitive market. A supplier's declaration is assessed by nobody — it is the supplier's claim about itself, which may perfectly well be true but is not independent testimony. All three arrive with a logo and a document number; only the first answers the question "can this laboratory measure this thing?"

What ISO/IEC 17025 actually requires

The 2017 edition is organised into five substantive parts: general requirements for impartiality and confidentiality, structural requirements, resource requirements (personnel, facilities, equipment, metrological traceability, externally provided products and services), process requirements, and management-system requirements ^[1]. The point that matters for anyone reading reports is that the standard does not prescribe which method is used. It prescribes the discipline around the method: that it will be appropriate, validated or verified, performed by authorised personnel, on equipment with traceable calibration, and that the result will carry its uncertainty.

That last row is the one most often broken and least often noticed. When a report says "complies" or "within specification", it has made a decision about a result that carried uncertainty — and the way that uncertainty was taken into account is called the decision rule. ILAC-G8 defines a decision rule as precisely that: a rule describing how measurement uncertainty is accounted for when stating conformity with a specified requirement, and ties it to the managed risk of false acceptance and false rejection ^[6]. Without a stated rule, a "pass" is an opinion with typographic confidence: a result marginally inside a limit may be outside it, and the reverse.

ESYD, ELOT, EOF: who does what in Greece

The architecture is European and deliberately monopolistic. Regulation (EC) No 765/2008 requires every Member State to appoint a single national accreditation body, which performs accreditation as a public authority activity, does not operate for profit and does not compete with the bodies it assesses; it is itself subject to peer evaluation, and only after passing it do its certificates enjoy a presumption of conformity across the Union ^[3]. In Greece that role belongs to ESYD, constituted as a non-profit legal entity of private law by Law 4468/2017 ^[4]. There is no second Greek accreditation body, and an "international accreditation certificate" issued by a private company is not accreditation in the legal sense of the word.

The two commonest confusions involve the other two bodies that appear in the same paperwork. ELOT is the national standards body: it publishes the Greek adoption of the standard, hence the title ELOT EN ISO/IEC 17025, but it accredits nobody. EOF is the competent authority for human medicines; it does not accredit testing laboratories, and its involvement in a material means the material has already moved into the medicinal field. Outward recognition runs through the multilateral arrangements: at the end of 2025 the ILAC MRA counted 121 signatories representing 122 economies, covering more than 114,600 accredited laboratories in total ^[9]. In practice a report from an accredited laboratory in another country is accepted here without repeating the test — provided, again, that the test was inside scope.

The scope of accreditation is the document that counts

The accreditation certificate is one page with a number on it. The scope of accreditation is the annex that accompanies it, often dozens of pages, and that is where the information lives: line by line, the material or class of materials, the measured property, the method with its code and usually the measuring range. That list is published by the accreditation body and kept current; it is not a document the laboratory supplies about itself. There is also a legitimate variant worth knowing about: the flexible scope, which lets an accredited body manage part of its own scope and modify or add activities without prior assessment of each change, within defined boundaries ^[7]. A flexible scope is not a relaxation, but it is a reason to ask for the current list rather than infer from an old printout.

• Certificate number and validity dates — accreditation expires, is renewed, suspended and reduced; the report date has to fall inside the period of validity.
• Material, not only method: accreditation for liquid chromatography in water or food does not cover a lyophilised synthetic substance — the same technique on another material is a different entry.
• Measured property: identity, purity and content are three separate lines, and a laboratory may be covered for one and not the other two; what an identity determination actually establishes is in any case narrower than the word suggests.
• Measuring range: a result outside the accredited range is outside scope, even with the right method on the right material.
• Fixed or flexible scope, and what the declared boundaries of the flexibility are ^[7].
• Subcontracting: where part of the testing is placed with another laboratory, the report has to say so, and the subcontractor's accreditation does not transfer to the issuing laboratory.

How to read a report carrying an accreditation symbol

An accredited laboratory can perfectly lawfully issue reports that fall outside its scope. Precisely because that is routine, the use of accreditation symbols and claims of accreditation status is regulated separately: ILAC-P8 sets supplementary requirements so that an accredited body does not, through its marking or its wording, create the impression that activities are covered when they are not ^[5]. The practical conclusion is unwelcome but simple: the logo at the top says nothing about the result line halfway down the page.

• The accreditation symbol always carries the certificate number of that specific laboratory; a logo without a number cannot be verified and does not count.
• Where covered and uncovered results appear in the same report, the report must distinguish them explicitly — if it does not, assume the worse case and ask.
• The method code on the report is matched line by line against the accreditation body's published list, not against the supplier's accompanying text.
• Every "within specification" judgement is accompanied by the decision rule that was applied ^[6].
• Opinions and interpretations are clearly separated from measured results; a remark such as "the product is of high quality" is not a test result ^[1].
• The sample identity as the laboratory recorded it matches the batch number on the label and on the delivery note.

There is one limit no accreditation crosses, and it deserves saying plainly. A scope of accreditation concerns the laboratory's competence, not the identity of the sample. An impeccably accredited laboratory measures accurately whatever is sent to it; who selected the sample, how it was drawn from the batch, and whether that batch is the same as yours all sit entirely outside the report. This is why matching the batch number across vial, test report and shipping document — the minimum record set that keeps material traceable — remains the only link that makes the report about the material in your hand, and it is a link no logo substitutes for.

Where the evidence is genuinely thin

Here we owe more honesty than the market usually offers. Validating an analytical method is not a property of the technique but of the combination of method, material and range: the Eurachem guide on the fitness for purpose of analytical methods insists that validation demonstrates suitability for a stated intended use, with the performance characteristics established in the material and over the range where the method will be applied ^[8]. For lyophilised synthetic peptides, such materials are rarely enumerated explicitly in published scopes of accreditation — and when they are absent, the liquid chromatography that produced the purity figure printed on the report "is accredited" without being accredited for this.

The second weakness is comparability. The standard requires a laboratory to monitor the validity of its results, and the principal way of doing so is participation in proficiency testing schemes, whose providers are governed by ISO/IEC 17043 ^[1][10]. For this class of material such schemes are scarce to non-existent; laboratories then meet the requirement through internal quality control, replicate testing and reference materials, which is lawful and reasonable but weak as evidence of agreement between laboratories. Nor is there any accredited certification scheme covering research peptides as a category: accreditation always concerns measurements, never products. The converse holds too — the absence of accreditation does not prove incompetence. A university laboratory may be technically excellent without ever having applied, because accreditation is expensive administrative infrastructure rather than a ranking of scientific merit.