How to Spot an Edited or Fake Certificate of Analysis

A certificate of analysis is a PDF, and a PDF can be edited. In a market where buyers choose suppliers largely on published lab results, altered and invented certificates circulate. This guide lists the checks that catch most of them, in the order worth doing.

1. Verify it at the laboratory first

Everything else on this page is a fallback. If the issuing laboratory publishes a lookup, use it: enter the report number and compare what the laboratory shows against what the vendor shows. A report that does not resolve at the lab, or resolves to different figures, ends the discussion. Our guide on verifying a Janoshik report walks through the fields to compare.

Be careful about where the verification link comes from. A link printed on a vendor’s page can point anywhere. Type the laboratory’s address yourself, then search for the report number there.

2. Check the lot number ties to the vial

A genuine certificate describes one lot. If the number on the label does not match the number on the report, the report may be real but it is not about your material — a common and deniable substitution. See lot numbers and traceability.

Watch for a report with no lot number at all, or one that says only “current batch”. That certificate cannot be tied to anything.

3. Look for internal inconsistency

• Dates that do not line up. An analysis dated before the lot was produced, or a report for a lot being sold a year later with no retest.
• Figures that contradict each other. Three vials reported at 10.2, 10.3 and 10.1 mg with a summary line saying 11 mg.
• A purity figure with no chromatogram, or a chromatogram whose visible peaks do not plausibly integrate to the stated percentage.
• Mass results that do not match the sequence. The reported molecular weight should match the compound, including any lipid chain. A mismatch of a few hundred daltons is not a rounding error.

4. Inspect the file itself

• Mixed fonts or misaligned text in the lot number, date or percentage fields — the exact places an editor would change. Zoom in; edited text often sits a pixel off the baseline or renders slightly differently from the rest of the line.
• A flattened image where the rest is selectable text. If the figures are a picture pasted over the original, that is deliberate.
• Document properties. Right-click → Document Properties shows the producing software and modification date. A laboratory report last modified in a photo editor is a finding.
• Screenshots instead of PDFs. A vendor who only ever shows cropped images of reports has made verification impossible on purpose.

5. Ask what is missing

Omission is more common than forgery, and harder to notice:

• Only the best of three vials shown, with the other two cropped out.
• A purity figure shown for one lot, with every later lot silently using the same certificate.
• An endotoxin result presented as though it proves sterility. It does not; see what an endotoxin figure means.
• A certificate for a blend showing one number, when a blend needs a figure per component. See how blends are tested.

6. Note who commissioned the test

The client named on the report is whoever paid for it. Often that is the manufacturer rather than the seller, which is normal — but it should be stated rather than hidden, and the seller should be able to explain the chain from that lot to the vial you receive. A vendor who presents a manufacturer’s report as its own in-house testing has already told you something about how it handles facts.

What good practice looks like

Figures published as reported, including the lowest of three vials. Lot numbers on the certificate and on the label. A report number that resolves at the laboratory. Plain statements where a lot was not tested for something, instead of a number borrowed from a different lot. That is the standard we hold ourselves to, and our results page is where it can be checked.

Materials supplied by BMS Labs are for laboratory research use only and are not for human or veterinary use.