How to verify a batch report before you buy
A certificate of analysis is a document about one sample from one lot. Here is every field on it, what that field establishes, what it leaves open, and what a measured result means when it comes back under the number on the label.
2 October 2026
Send me the checklist
One email when a new hub goes up. Research use only; nothing else, and one click to leave.
Every vial on this bench comes with a page, and most of those pages carry a number in the title position: 10 mg, 99% pure, pharmaceutical grade. None of those is a measurement. A measurement looks different — it has a method beside it, a laboratory above it, a lot number on it and a date on it. This page is about the second kind, and specifically about the gap between holding one of the second kind and knowing what it describes.
The gap is not a technicality. The Belgian public-health laboratory screened the ten most frequently encountered falsified peptide drugs on its market, bought from three different suspected illegal internet pharmacies, and found purity between 5% and 75% for the cysteine-containing peptides, with arsenic and lead at up to ten times the ICH toxicity limit for parenteral drugs. Every one of those vials arrived in packaging that described itself. The Hungarian surveillance of online semaglutide sellers is the same story with numbers attached: three vials delivered, three judged substandard and falsified on visual inspection, and a measured purity of 7.7% to 14.37% against a 99% figure printed on the label.
What a certificate of analysis actually is
A certificate of analysis is a record of measurements performed on a sample, printed by the body that performed them. That is the whole of it. It is not a grade and it is not a pass mark, because the word “certificate” describes the document’s type and not its verdict.
The distinction that catches people is between a specification and a result. ICH Q6A is the guideline that defines what a specification is: a set of test procedures together with the acceptance criteria a result has to meet. A specification is a decision made in advance about what to measure and what counts as acceptable. A certificate of analysis is what one sample returned when somebody measured it against that decision. Both documents matter and neither substitutes for the other, and a certificate that lists results without naming the methods underneath them has handed you half the record.
Which brings up the panel structure. The Belgian screening ran five: the active ingredient itself, impurities related to the active ingredient, small-molecule contaminants, elemental impurities and residual solvents. Any one of those five can be published on its own and still be a true statement about the vial. A headline purity figure is one of the five with the other four removed, which is why the word “pure” on a product page is worth less than the sentence naming which measurement produced it.
The seven checks, in the order they catch things
The lot number. Does the number printed on the report match the number printed on your carton or label? This is the least effort on the list and the one most often skipped, and it is the one that decides whether the rest of the document is about your object at all. A genuine report for a different lot is still a genuine report, and it still tells you nothing about what arrived. The mapping from report to carton is made by the seller, not by the laboratory.
The analysis date against the manufacture date. A report is a photograph, not a label. It records what was in the sampled material on the day the measurement was taken, and the material keeps degrading whether or not anyone is watching. A value does not travel forward with the vial.
The laboratory’s name and scope. A name and an address on the letterhead is what makes a document traceable rather than anonymous. It is an attribution and nothing more. The follow-up question is the one that does the work: what is this laboratory accredited for, and is this particular method inside that scope? Accreditation attaches to a scope, not to a building.
The accreditation mark, for what it is worth. In the European Union the framework that accreditation bodies and conformity assessment bodies operate under is Regulation (EC) No 765/2008. Worth being blunt about what a mark on a report means: it is a claim about the laboratory and never a claim about the product. No accreditation mark anywhere certifies a vial.
The sample. How many units, drawn how, from where in the lot? If the report does not say, then “tested” on a product page describes a test somebody ran once, with no plan behind it. The Hungarian team bought three injection vials from the illegal pharmacies with the most links and the lowest prices; none of the three Ozempic prefilled pens they ordered arrived at all, which is the other outcome of buying a medicine from an unlicensed website. The three vials that did arrive failed visual inspection against more than half — 59% to 63% — of the criteria applied.
The method names. A value with no method beside it is a number with nothing behind it, and a method is built around a target list. A clean result is clean against that list.
The number against the claim. This is the check that pays for the rest. The semaglutide vials exceeded their labelled content by 28.56% to 38.69% while measuring 7.7% to 14.37% purity against a 99% claim. Those are not two separate defects — that is one vial that is over-concentrated and impure at the same time, and no single headline number describes it. The same study found no peptide-like impurities and no viable microorganisms, and endotoxin in every single sample at 2.1645 to 8.9511 EU/mg. A good result on most panels and a positive on the one that matters for an injectable is the ordinary shape of this problem, not an unusual one.
Why one sample is not a claim about your vial
A vial is a manufactured dose. Even in a GMP-controlled process, a result on a sample is a result on a sample, and the sampling plan — not the assay — is what connects the two. In the semaglutide work the buyers received three vials and each was individually non-compliant, which is a more useful finding than a per-lot percentage would have been, because it is the only form in which the answer is useful to the person holding the parcel.
Two more things a well-formed report still does not establish, and both were on the record before this question was asked. It does not extend past the expiry printed on the paper. And it does not say anything about what happened to the vial between the laboratory and your hands.
What we publish, and what we do not
Where a maker publishes a report for a batch, it is already on the product page under Batch documentation — the laboratory, the batch, the measured values and the verification key, as its issuer printed them. We add nothing to it and we correct nothing in it. Where no maker publishes one, the line on the product says that in plain words rather than sitting there as a card that looks like it failed to load.
We do not publish an assay, a purity figure, a residual-solvent result or a mass balance for any line ourselves, and we will not print a number we cannot trace to a document. Every report on this site describes the batch printed on it, which may not be the carton you receive; send your order number and we request the paperwork for the batch that actually shipped.
What the tests themselves prove, and what each one leaves open, is on how to read a supplier documentation claim — that page works through the six tests on a manufactured dose, this one works through the document that reports them. The regulatory file on a licensed oral androgen, and what an approval rests on, is in what the record says about oral androgens.
We write no dosing protocols, we cannot promise a result, and nothing on this page is medical advice.
⚠ Everything sold here is for in-vitro laboratory research only. Not for human or veterinary use.
Field by field: what each part of a batch report carries, and the question it does not answer
| What | What the field on the report is | What it establishes | What it leaves open, and the line it does not license | What it means for you |
|---|---|---|---|---|
| "Certificate of analysis" — the document's own name and type | That a named body performed named test procedures on a sample and printed the values it read | That the sample was the product. The document type is a statement about a test, not a grade, and a certificate can be issued for a falsified product just as easily as for a genuine one | Nothing on the face of the word "certificate" carries a quality judgement. ICH Q6A is where a specification — the set of tests with their acceptance criteria — is defined, and a specification is a decision about what to test, not a measurement of what came back | The document's title is the least informative line on it. Read the values and the method names, and ignore the heading. |
| The lot or batch number on the report | Which batch the sample was drawn from, as the issuing body recorded it | That the carton in your hand came out of that batch. The mapping from report to carton is made by the seller, not by the laboratory, and it is the single easiest thing to get wrong | A report and a lot number that agree with each other say nothing about the lot number printed on the box until you have put the two side by side and seen they match | This is the first comparison to make and the one most often skipped. A genuine report for a different lot is still a genuine report, and it still tells you nothing about what you received. |
| The date of analysis and the manufacture or expiry date on the label | When the measurement was taken, and what the maker states about the lot's shelf period | That the number describes material in the state it was in on that date. Peptides continue to degrade whether or not a report exists, and a value does not travel forward with the vial | A report dated months before the date on your carton is describing something earlier in the lot's life. Nothing in the document says the material has not changed since | Read the analysis date as a photograph, not as a label. What the report cannot tell you is whether the vial has been kept since. |
| The name, address and scope of the laboratory | Who ran the work and where, which is what makes the document traceable rather than anonymous | That the laboratory is competent to run the test it ran. A name on letterhead is an attribution; it is not evidence of oversight | "Analysed by" is not a claim about scope. Ask what the laboratory is accredited for, and for which specific test, before you read anything else on the page | A named laboratory is the difference between a document and a screenshot. An unnameable one is worth exactly as much as the image it came from. |
| An accreditation mark, and the standard behind it | That a third party has assessed the laboratory against a published competence standard for a stated scope | That every test the laboratory runs is covered. Accreditation is granted against a scope, and a report for a method outside that scope carries the mark but not the assurance | An accreditation mark on a report is a claim about the laboratory, never about the product. No accreditation mark anywhere certifies a vial | In the European Union the framework that accreditation bodies and conformity assessment operate under is Regulation (EC) No 765/2008, which is the standard a mark on the page is measured against. |
| The sample: how many units, and which ones | The size and the drawing plan of what was tested, if the report states them | That your unit resembles the units that were tested. One sample from one lot is a statement about the sample. A vial is a manufactured dose, and a dose is an individual object | "Tested" on a product page, with no sample size, no method and no plan, describes a test somebody ran once. It is not a statement about the flask in the parcel | The Hungarian semaglutide surveillance bought three injection vials and judged all three substandard and falsified on visual inspection alone, with non-compliance in 59–63% of the criteria it applied. Packaging and labelling are the easiest part of a forgery to get wrong. |
| The method name printed next to each value | That the number came from a stated procedure, so it can in principle be repeated | That the procedure detected what it was designed to detect. A method is built around a target list, and a clean report is clean against that list rather than against every possible species | A value with no method beside it is a number with nothing behind it. The Belgian peptide screening measured API, API-related impurities, small-molecule contaminants, elemental impurities and residual solvents as five separate panels, and the headline is one of the five | Ask for the method before you read the result. "Purity" without a method is a phrase, not a measurement. |
| The measured assay value against the figure on the label | How much of the named substance was in the units sampled, on the day they were sampled | That the material is fit for anything, or that the number is a grade. Assay is one row of a specification, and the rest of the specification is not implied by it | Nothing that reconciles the measured number with the declared one. In the semaglutide surveillance the measured content exceeded the labelled amount by 28.56–38.69% while the measured purity came back at 7.7–14.37% against a 99% claim on the label — both directions at once, in the same vials | When the measured figure sits under the claim, the claim is what is wrong. That is not a rounding difference to be explained away; it is the measurement doing the only job it has. |
| The panels that were not run | Nothing. An absent test is not a passing result, and the gap is usually the informative part | Everything that panel would have covered. In the same semaglutide vials, no peptide-like impurities were identified and no viable microorganisms were present, and endotoxin was nonetheless detected in every sample at 2.1645 to 8.9511 EU/mg | A clean sheet on the tests that were run says nothing about the ones that were not. In the Belgian screen, arsenic and lead reached up to ten times the ICH toxicity limit for parenteral drugs, and speciation confirmed all the arsenic was in the more toxic inorganic form | Read a report for what is missing as hard as for what is printed. The endotoxin panel above is the clearest example in the published record: a good result on every other panel, and a positive on the one that matters for an injectable. |
Questions this page answers.
So is a certificate of analysis worthless?
No, and it is not what it is usually sold as either. A certificate of analysis is a record of measurements made on a sample, and that is genuinely useful — it is how you learn that a number you were quoted is not the number on the label. What it is not is a pass mark, a grade, or a guarantee about the object that arrives in the box. A document can be authentic, current, correctly signed and still describe a different lot from the one in your carton.
How do I check that a laboratory is credible?
Start with the name and the address on the report, then ask what the laboratory is accredited for and for which specific method. Accreditation is granted against a scope rather than against a laboratory as a whole, so the useful question is not whether a mark is on the page but whether this test falls inside it. In the European Union the framework those accreditation bodies and conformity assessment bodies operate under is Regulation (EC) No 765/2008. A laboratory that cannot be named, or that answers nothing about its scope, has told you the thing you needed to know.
What does it mean when even the measured purity is below what was claimed?
It means the claim was wrong, and the measurement is the more reliable of the two. The semaglutide vials bought from illegal online pharmacies in the 2024 Hungarian surveillance measured between 7.7% and 14.37% purity against a 99% figure on the label, and their content ran 28.56–38.69% above the labelled amount — so the same vials were simultaneously over-concentrated and impure. On the Belgian market, the ten most frequently encountered falsified peptide drugs screened at 5% to 75% purity for cysteine-containing peptides. A number that comes back under the claim is not a defect to be argued with; it is the measurement doing its job.
What does this shop publish, and what does it not?
Where a maker publishes a report for a batch, it is already on the product page under Batch documentation, with the laboratory, the batch, the measured values and the verification key as its issuer printed them, and we add nothing to it. Where no maker publishes one, we say that in plain words on the line rather than leaving a card that looks empty. We do not publish an assay, a purity figure, a residual-solvent result or a mass balance for any line ourselves, and we do not paraphrase a supplier's handling instruction as though it were a measured result. Every report on this site covers the batch printed on it, which may not be the carton you receive — send your order number and we request the paperwork for the batch that actually shipped.
Where this comes from.
- Ashraf AR, Mackey TK, Vida RG, et al. Multifactor Quality and Safety Analysis of Semaglutide Products Sold by Online Sellers Without a Prescription: Market Surveillance, Content Analysis, and Product Purchase Evaluation Study. J Med Internet Res 2024;26:e65440. (PMID 39509151) — 1,080 links collected from search engine results pages; 317 (29.35%) were online pharmacies, of which 134 (42.3%) led to 59 unique illegal online pharmacy websites; the top 30 domains accumulated over 4.7 million visits between July and September 2023; three injection vial purchases were delivered and none of three Ozempic prefilled pens arrived; all three vials were considered probable substandard and falsified with visual non-compliance in 59–63% of evaluated criteria; semaglutide content exceeded labelled amounts by 28.56–38.69%; no peptide-like impurities were identified and the lyophilised samples were devoid of viable microorganisms at the time of testing; endotoxin was detected in all samples at 2.1645–8.9511 EU/mg; measured purity ranged 7.7–14.37% against the 99% claimed on the labels.
- Janvier S, Cheyns K, Canfyn M, et al. Impurity profiling of the most frequently encountered falsified polypeptide drugs on the Belgian market. Talanta 2018;188:795-807. (PMID 30029448) — a systematic screening of the ten most frequently encountered falsified peptide drugs on the Belgian market, acquired from three different suspected illegal internet pharmacies, analysing the active ingredient, API-related impurities, small-molecule contaminants, elemental impurities and residual solvents; purity ranged between 5% and 75% for cysteine-containing peptides; one sample was contaminated with lead and multiple samples carried arsenic and lead at up to ten times the ICH toxicity limit for parenteral drugs; arsenic speciation confirmed all arsenic was present in the more toxic inorganic form.
- ICH Q6A — Specifications: Test Procedures and Acceptance Criteria for Active Substances. The guideline in which a specification is defined: the set of tests to be performed and the criteria a result must meet. A specification is what to measure and what counts as acceptable; a certificate of analysis is the values one sample returned against that specification.
- ICH Q7 — Good Manufacturing Practice for Active Substances. The guideline that sets out what a manufacturing quality system is expected to cover, and therefore what a document emerging from one is and is not evidence of.
- Regulation (EC) No 765/2008 on the accreditation and market surveillance relating to the marketing of products, and repealing Directive 2001/16/EC — the European Union framework under which accreditation bodies and conformity assessment bodies operate, and therefore the framework an accreditation mark on a report is measured against. EUR-Lex.
- Regulation (EU) No 536/2014 on clinical trials on medicinal products for human use, Article 2(1) — which applies by reference the definition of "medicinal product" in Article 1(2)(b) of Directive 2001/83/EC, and defines an investigational medicinal product in Article 2(1)(5) as a medicinal product being tested or used as a reference in a clinical trial. A molecule outside that frame has no authorised route to a customer, which is why the paperwork on a research vial and the paperwork on a licensed medicine are not the same paperwork. EUR-Lex.
