Peptifact

"Third-Party Tested": What the Claim Commits a Seller To, and the Four Questions It Leaves Open

Third-party testing is a claim about who ran an assay, not about what the result was, which lot it came from, or whether the laboratory was competent to run it. Each of those is separately checkable, and one of them is checkable for free in a public register.

Robert F · Edited by Caroline S · Published 2026-09-14

Illustration: An open blank register book and a magnifying glass on a grey desk.
Illustration

"Third-party tested" is the most widely printed quality claim in this market and the least specific. It appears on storefronts, on product pages and in vendor comparisons, usually with no laboratory named, no assay named and no certificate attached.

The phrase is not empty. It commits a seller to a checkable fact: somebody other than the seller ran an assay. That is a real claim and a real improvement on nothing.

What it does not do is answer the four questions that decide whether the claim tells a buyer anything about the vial in their hand. This page sets out what the phrase asserts, what it leaves open, and which of the gaps can be closed for free by a reader with ten minutes.

What the phrase asserts

One thing: that the party running the assay was not the party selling the product.

That is worth having. A seller's in-house purity figure and an outside laboratory's purity figure are different kinds of evidence, and the second is harder to arrange to taste.

It is also the whole of the claim. The phrase contains no assertion about:

  • which test was run — a mass-spectrometry identity check and an HPLC purity assay answer different questions, and neither answers the other, as the assay breakdown sets out;
  • what the result was — "tested" is not "passed", and a claim that stops at the verb has not reported an outcome;
  • which lot was tested, or whether it is the lot that shipped;
  • who selected the sample;
  • whether the laboratory was competent for that test on that material.

Each gap is separately closable. Two of them are closable by the reader alone.

Accreditation is not certification, and the words are not decorative

The most common confusion in this area is between two things that sound alike.

Certification applies to products, processes and management systems, and is performed by certification bodies.

Accreditation is the formal recognition, by an accreditation body, that a conformity assessment body — a testing laboratory, for example — is competent to carry out specific activities. For testing laboratories the standard is ISO/IEC 17025.

The distinction is maintained structurally at the international level. ILAC, the International Laboratory Accreditation Cooperation, coordinates laboratory accreditation and states in its own FAQ that it "is not involved in the certification process", which sits with a separate organisation, IAF.

So a laboratory that advertises itself as "ISO/IEC-certified" — and this wording does appear on the homepages of laboratories serving this market — has described its own status with the wrong word. That is not proof of anything untoward. A small laboratory writing its own marketing copy can get a term of art wrong. But the wording is a signal about how carefully the claim was made, and it is visible from the outside in five seconds.

Accreditation has a scope, and the scope is the substance

Here is the part that matters most and gets mentioned least.

Accreditation is not granted to a laboratory as an institution. It is granted for a scope: a defined list of test methods, applied to defined types of material. The scope document is published alongside the certificate.

The practical consequence: a laboratory can hold entirely genuine ISO/IEC 17025 accreditation and still have run the test on your certificate outside it.

Take a real, ordinary example. One analytical laboratory whose name appears on certificates circulating in this market describes its business, on its own homepage, as serving the dietary supplements, food, life science and pharmaceutical, and cosmetics industries — listing raw materials analysis, assays, potency and purity, microbiology, heavy metals, Brix determination, protein content and kava testing. Nothing on that page presents it as a research-peptide laboratory, and nothing on it is discreditable. It is a contract laboratory doing contract laboratory work.

But a scope covering heavy metals in dietary supplements is not a scope covering purity determination of a synthetic peptide. The accreditation would be real, and it would not reach the test.

Checking this needs three items:

  1. The accreditation body's name — A2LA, ANAB, PJLA, IAS, or a national body outside the US.
  2. The certificate number.
  3. The scope document, and a look for the specific method and material.

Then look it up. ILAC's own answer on verification is that "the accreditation bodies are responsible for maintaining the list of laboratories and inspection bodies that they have accredited", and that an enquirer should consult "the accreditation body's online directory of accredited laboratories and inspection bodies on their website".

Those directories are free. This is the one link in the chain a buyer can verify at no cost, from outside, in a few minutes — and a claim that arrives without a body and a number cannot be looked up at all.

Who chose the sample

A laboratory tests what arrives in the post.

If the seller selected the vial, packed it and sent it, then the independence in "third-party tested" attaches to the analysis and not to the selection. That is not an allegation. It is the ordinary shape of the arrangement, and it is the reason a testing programme that buys its own samples anonymously generates a different class of evidence.

It is also why this publication's testing standard specifies anonymous purchases placed weeks apart rather than samples supplied by a vendor. A supplied sample can be a perfectly honest sample. It just cannot demonstrate, by itself, that it is representative of what a customer receives.

The lot number is the whole question

Of the four gaps, this is the one that decides whether any of it applies to the object on the desk.

A certificate names the batch that was sampled. A vial carries a batch number. Either they match or they do not, and the comparison takes ten seconds.

Three outcomes, and they are not equally bad:

  • They match. The certificate is about the material in hand, subject to everything else on this page.
  • They differ. The certificate is about a different batch. It may still be informative about the manufacturer; it is not evidence about this vial.
  • One or both are absent. Nothing can be determined, and the claim has no anchor at all. This is the most common case and the worst of the three, because it looks like the first.

There is a further step that is often taken silently. A laboratory result is normally a statement about the sample received. Extending it to a production run is an inference — a reasonable one where sampling was designed for it, and an unsupported one where a single vial was posted in. Whoever publishes the certificate makes that extension; the laboratory generally does not.

The fields a certificate carries, and the four things it cannot establish, are set out on how to read a certificate of analysis.

A useful parallel from a different regulation

FDA has considered, in a neighbouring context, whether a self-certifying mechanism shifts responsibility to the customer.

In its 2013 final guidance on research-use labelling, addressing storefronts that gate a purchase behind a user certification, the agency wrote that such a programme "would be viewed as one factor to consider" — and then that "the existence of a certification program alone would not relieve manufacturers from their responsibilities to ensure that their labeling and distribution practices... are consistent with the product's RUO/IUO label."

That is about labelling rather than testing, and it is explicitly non-binding. The transferable idea is the allocation: a document produced to demonstrate compliance is evidence to be weighed, not a discharge of the underlying obligation. A certificate on a product page works the same way. It is one factor. It is not the answer to whether the material is what the page says it is.

The regulation itself, and what its label does and does not cover, is on "research use only": two regulations, two labels.

The five questions, in order

Put to any page carrying the phrase:

  1. Which laboratory? A name, not "an independent lab".
  2. Accredited by whom, under what number, for what scope? Then check the body's public directory, and check the scope reaches the method and the material.
  3. Which assay, and what was the result? Tested is not passed.
  4. Which lot, and does it match the vial?
  5. Who selected and submitted the sample?

A seller answering all five has made a claim worth something. A seller answering the first two has made it checkable. A seller answering none has printed a phrase.

What this page does not say

It does not say that third-party testing is worthless — it is the single most useful thing a seller in this market can arrange, and a page criticising the phrase should not be read as criticising the practice.

It does not say that any named laboratory is incompetent, unaccredited, or wrong. The Colmaric example above is used for one narrow and checkable point — its own homepage describes a scope of industries that does not include research peptides, and uses "ISO/IEC-certified" where the term of art is accredited. No claim is made here about its accreditation status, which we did not attempt to determine, or about the quality of its work.

And it does not tell a reader that an unverifiable claim is a false one. Most of them are probably true. The point is narrower and, for a buyer, more useful: a claim nobody can check is not evidence, whatever its truth.

Sources and dates

  • ILAC (International Laboratory Accreditation Cooperation), FAQs. Statements on the ILAC/IAF division of responsibility, accreditation bodies' maintenance of accredited-laboratory lists, and the direction to consult an accreditation body's online directory. Read 14 September 2026.
  • ISO/IEC 17025, General requirements for the competence of testing and calibration laboratories — referenced as the standard against which testing laboratories are accredited.
  • Colmaric Analyticals, company homepage. Industries served, listed capabilities, and the phrase "Our ISO/IEC-certified testing lab". Read 14 September 2026.
  • FDA, Distribution of In Vitro Diagnostic Products Labeled for Research Use Only or Investigational Use Only: Guidance for Industry and Food and Drug Administration Staff. Final guidance, November 2013, CDRH and CBER. Section IV.C.1, "Use of a 'certification program'". Read 14 September 2026.

Frequently asked questions

What does third-party tested actually mean?

Literally, that an assay was run by someone other than the seller and the manufacturer. That is all it asserts. It does not name the test, state the result, identify the lot, say who selected the sample, or claim the laboratory was competent to perform that test on that kind of material. Each of those is a separate question with a separate answer, and a page carrying the phrase and none of the answers has made the weakest version of the claim available.

Is a third-party lab the same as an accredited lab?

No. Third-party describes a relationship — the laboratory is not the seller. Accredited describes an assessed competence — an accreditation body has evaluated the laboratory against ISO/IEC 17025 and recognised it as competent for specific tests. A laboratory can be a genuine third party and hold no accreditation at all. The two claims are often printed together as though one implied the other; neither implies the other.

What is the difference between accreditation and certification?

Certification applies to products, processes and management systems, and is carried out by certification bodies. Accreditation is the formal recognition by an accreditation body that a conformity assessment body — a testing laboratory, for instance — is competent to carry out specific activities. The distinction is maintained at the international level: ILAC coordinates laboratory accreditation and states that it "is not involved in the certification process", which sits with IAF. A laboratory advertising itself as "ISO/IEC-certified" has described its status with the wrong word.

How do I check whether a laboratory is really accredited?

Ask the seller or the laboratory for three things: the name of the accreditation body, the certificate number, and the scope of accreditation. Then look the certificate up in that body's public directory. ILAC's own guidance is that accreditation bodies maintain the lists and that an enquirer should check "the accreditation body's online directory of accredited laboratories and inspection bodies on their website". If the claim comes without a body and a number, there is nothing to look up — and that absence is itself informative.

Why does the scope of accreditation matter?

Because accreditation is granted test by test, not to the laboratory as an institution. A scope document lists the specific methods a laboratory is accredited for and the materials it is accredited to run them on. A laboratory accredited for heavy-metals testing in dietary supplements is genuinely accredited — and that accreditation says nothing about its competence to run a purity assay on a synthetic peptide. Checking the scope is the step that turns a logo into evidence, and it is the step almost nobody takes.

Does a third-party result apply to the vial I received?

Only if the lot numbers match, and the phrase itself makes no such claim. A certificate names the batch that was sampled. A shipment arrives with a batch number on the vial. Those are two facts that either agree or do not, and comparing them takes ten seconds. Where a certificate carries no lot number, or the vial carries none, the question cannot be answered at all — which is a different situation from a mismatch, and worse.

Who picked the sample the laboratory tested?

Usually the seller, and the claim does not say otherwise. A laboratory tests what arrives in the post; independence of the analysis is not independence of the sampling. This is not an accusation about any particular seller — it is the ordinary structure of the arrangement, and it is why a testing programme that selects and buys its own samples anonymously produces a different kind of evidence from one that receives them. It is also why this publication's own testing standard specifies anonymous purchases rather than supplied samples.