How to verify any COA in 2 minutes

Why a routine rather than a long checklist?

The checks below are ordered by how much they rule out for the least effort, with one rule: stop at the first step that fails. If the batch does not match your vial, nothing further on the page can tell you about your vial. The timings are a guide: a report with a working verification route takes about two minutes, and one without takes longer, which is itself information.

Step 1: Does it describe my vial? (about 20 seconds)

• Compound. The exact name, not a family. A report for semaglutide does not describe a vial labelled tirzepatide.
• Label amount. A report for a 10 mg vial does not describe a 20 mg vial.
• Batch or lot code, character by character. ICH Q7 defines a batch number as the code "from which the production and distribution history can be determined" (ICH Q7). A different code is a different history.

If any of the three fails, or the report names no batch, stop. The document may be genuine, but it is not about your vial (lesson 11).

Step 2: What was actually tested? (about 20 seconds)

Before reading any number, list the tests the report covers. Tick off identity (a mass, from a named method such as LC-MS), purity (HPLC), amount against the label, and endotoxin. Note anything else, such as water or residual solvents, and note what is missing.

A purity-only report has not checked which molecule is in the vial. An endotoxin-only report has checked neither identity nor purity. Write the scope down so you do not fill in tests that were not run.

Step 3: Does the issuer confirm it? (about 40 seconds)

This is the check that catches an edited or invented document (lesson 14). Do it in this order:

1. Read the issuer's name off the report.
2. Find the issuer's website yourself, by search or by typing the address. Do not use a link, QR code or domain the seller gave you.
3. Find the verification route on that site and enter the report's codes exactly as printed.
4. Compare the record line by line with your copy: compound, label amount, batch, dates and every result.

If the record matches, the document is genuine and unaltered. If anything differs, trust the issuer's record. If there is no verification route at all, the report is unverifiable by you: not fake, but not evidence either. For a worked example with one widely used issuer, see how to verify a Janoshik report.

If the issuer publishes a file fingerprint, you can also compare the SHA-256 of your copy. Under the US federal hash standard, such digests "are used to detect whether messages have been changed since the digests were generated" (NIST, FIPS 180-4). It is optional, but conclusive when it matches.

Step 4: Who signed it, and can they be checked? (about 20 seconds)

For a report issued in the Philippines, Section 40 of the Chemistry Profession Act provides that a chemistry document is not official "unless such document is signed by a registered chemist and the chemists registration number affixed thereto" (RA 10657, Lawphil). Enter the name or number on the PRC verification service (PRC LERIS).

For a report from abroad, the rules of that country apply. If the report claims accreditation, note the accreditation body and certificate number. ILAC describes its members as accreditation bodies that accredit testing laboratories against ISO/IEC 17025, and its site carries a search of the accreditation bodies it recognises (ILAC), where the claimed body can be checked before you look up the certificate on that body's own register. Accreditation is not required for a report to be genuine, but a claimed accreditation that cannot be found is a problem.

Step 5: Do the numbers make sense? (about 20 seconds)

• Dates in order. Received before analysed, analysed before reported, and none of them before the batch could have existed.
• Expected mass fits the compound. Compare the expected mass on the report with the published value on the compound page, such as tirzepatide. The mass checker does the charge-state arithmetic from lesson 7.
• Purity matches its peak areas, if areas are given. Main peak area divided by total area should give the stated figure.
• Amount is plausible for the vial. A measured amount in the wrong units, or wildly beyond the label, points to a transcription error or worse.

These checks catch honest slips as well as edits; lesson 13 shows what each looks like.

What does each outcome mean?

What does a worked example look like?

An illustrative case, with invented details. A buyer holds a 10 mg tirzepatide vial with batch code TZ-0415. The seller sends a certificate from an overseas laboratory showing tirzepatide, 10 mg, batch TZ-0412, purity 99.3%, analysed four months before the purchase.

1. Step 1: compound and size match, batch does not. 0412 is not 0415.
2. The routine stops. Whatever the purity says, it describes a different batch.
3. The buyer asks for the TZ-0415 report. If there is none, this batch has not been shown to be analysed.

The COA checker runs the same questions interactively and notes where you stopped.

Check your understanding