Certificate of Analysis, Conformance and Origin: Three Different Documents
Open a document headed “Certificate of Analysis” and run your eye down the column where the results should be. If every single row reads “Conforms”, or “Pass”, or “Complies”, then whatever the heading claims, no measurement has been disclosed to you. The document is a conformance statement in an analysis costume. It is not necessarily dishonest, but a reader who mistakes it for a data report will draw conclusions the paper never underwrote.
A handful of verdict rows is ordinary enough. Certain tests really do yield a judgment rather than a figure, and a visual appearance check is the obvious example. It is the all-verdict document that deserves a second look.
Sorting them in ten seconds
Three kinds of paper travel with materials, all abbreviated to something starting with “certificate”, and only one of them contains measurements. The differences are not matters of house style; they concern what is being asserted and who is doing the asserting. A quick triage:
- Every attribute paired with a value and a named method? An analysis.
- Attributes paired with verdicts instead? A conformance statement, regardless of the title at the top.
- A country named and no attributes at all? A trade document.
The document that carries measurements
Results are what a certificate of analysis exists to report. Its signature characteristic is that every attribute arrives in three parts: the value obtained, the method that obtained it, and the criterion it was tested against. That construction is explained in how a specification limit gets set.
Because the results came from one particular lot, the document belongs to that lot and no other. In principle it is reproducible, meaning a second laboratory running the same method on the same material ought to arrive somewhere close, within the spread described under measurement uncertainty. Most importantly it is falsifiable, and that is the source of its authority. A checkable claim differs in kind from an uncheckable one.
The document that carries a conclusion
Certificates of conformance, sometimes titled conformity or compliance, assert that a lot satisfies its specification. Expect a lot number, a reference to the specification, a statement of conformity and a signature. Data is the one thing absent.
Describing this as deception would be wrong. It is a different instrument with legitimate uses: transmitting a conclusion while withholding the underlying results suits situations where a method is proprietary, or where a distributor is relaying an assurance it received from further upstream.
What follows from that is the real point. Evaluation is impossible. There is no chromatogram to examine, no integration to interrogate in the manner of peak integration, and no figure to set beside another lot. Such a document asks for belief, and everything it is worth depends on the credibility of the signatory.
The document that carries a country
Where goods were manufactured is all a certificate of origin states. It belongs to customs and trade rather than to quality, and the issuing or endorsing body is typically a chamber of commerce or an equivalent organization rather than any laboratory.
Composition, purity, whether testing occurred at all: none of this is addressed. For a buyer its relevance runs to duty treatment and import paperwork, alongside the matters raised in importing research chemicals into the US. Anyone presenting one as evidence of quality has made a category error, and sometimes a calculated one.
Signatures and what stands behind them
Format matters less than issuer. Four arrangements can look nearly identical on the page while differing considerably in weight.
| Issued by the testing laboratory | Whoever signed also did the work. The strongest of the four, and the scope question discussed under ISO/IEC 17025 accreditation bears on it. |
| Issued by the manufacturer from its own testing | Both common and legitimate. Incentives here are examined in third-party versus in-house testing. |
| Reissued by a distributor on its own letterhead | The chain becomes the issue. Lot identity has to tie the paper to the container you are holding, which is what batch and lot numbering is for. |
| Distributor conformance statement based on an unverified upstream certificate | An assertion resting on somebody else’s assertion. Weakest of the four by a clear margin. |
The date stamp nobody reads
Every one of these three documents describes a lot as it stood at the moment of assessment. Not one of them describes what became of the container afterwards: the conditions under which it traveled, the temperature excursions covered in cold chain and excursions, how it has been stored since arrival, or how many times the vial has been opened since.
Think of any certificate as a photograph with a date written on it. Reading it as a description of whatever currently sits in the freezer is the most frequent mistake made with all three document types, and the one whose consequences reach furthest.
Before you file it
Locate the lot number and confirm it matches the container in front of you. Paper that does not connect to the physical material is not evidence about that material, whichever of the three kinds it happens to be. Preserving that connection is exactly the purpose of chain of custody.
