How to read a peptide certificate of analysis (COA)

A certificate of analysis is a lab report for one lot. Used well, it answers a narrow question: did this sample match the expected molecule, and how much else was in the vial? Used poorly, a purity percentage is treated as a safety rating. It isn’t.

This is a reading guide, not a buying guide. A COA does not make a research chemical into a medicine. See what the labels mean.

1. Match the paper to the vial

Before any chromatography, check the boring lines:

• Name and sequence should match the peptide you think you are looking at — including caps, amidation, disulfide bridges, or a metal (GHK vs GHK-Cu).
• Lot or batch number on the COA must match the vial. A report for lot A says nothing about lot B.
• Date of analysis and any retest or expiry date. Peptides can degrade with heat, moisture and time.
• Amount tested vs labeled contents. A 5 mg vial with a COA run on a different fill size is a mismatch.

If those fields are blank, stock, or clearly copied from another product, stop there. The rest of the document is unmoored.

2. Identity — is it the right molecule?

Purity without identity is a clean measurement of the wrong thing. Look for a method that can actually distinguish the sequence:

• Mass spectrometry (MS) should report a mass that matches the calculated mass of that sequence (and the expected salt or adducts). A gap of many daltons is a red flag.
• HPLC retention time compared with a reference standard is supporting evidence, not a fingerprint by itself.
• Amino-acid analysis or MS/MS sequence confirmation is stronger, and often missing on marketing PDFs.

If the only “identity” test is appearance (“white powder”) or a catalog number, that is not identity testing.

3. Purity — what “99% HPLC” actually means

HPLC purity is usually peak-area percent: how large the main peak is compared with other peaks the detector can see, under those conditions. It is not:

• percent by weight of the vial
• proof there are no salts, water, or solvents (those often don’t show up as peptide peaks)
• proof of safety or biological activity

A 99% HPLC result can sit next to a large trifluoroacetate (TFA) or water fraction. TFA is a leftover from typical peptide purification. That is why serious reports also give net peptide content (how much actual peptide is in the powder) separately from HPLC purity.

Ask whether the chromatogram is attached. A number without a trace is easier to fake or recycle.

4. Quantity — labeled milligrams vs what was measured

Look for a quantitative method (HPLC against a standard, or amino-acid analysis) and a net peptide content or assay value. “10 mg” on a cap is a fill claim. The COA should say how much peptide was found in the sample that was tested.

If quantity is missing, you have a qualitative story, not a dose of anything — and Peptide World does not discuss dosing.

5. Leftovers the purity peak can hide

Useful COAs call out what HPLC area-% can miss:

• Counterions such as TFA or acetate
• Water / loss on drying
• Residual solvents
• Heavy metals (less common, still relevant for some routes of synthesis)

None of these lines turn a research chemical into a drug. They tell you whether the powder is mostly peptide or mostly something else.

6. What is often missing — and why that matters

Many research-chemical PDFs stop at identity and HPLC purity. For anything intended as a sterile drug, quality systems also look at:

• Bacterial endotoxin (USP 〈85〉 is the usual pharmaceutical test)
• Sterility
• Particulates
• Validated methods (see ICH Q2) and a lab that is competent to run them

Absence of those tests is common on “research use only” paperwork. It is also the point: a COA is not a sterile-product release. The FDA is explicit that compounded drugs are not FDA-approved and that insanitary conditions are a separate, serious problem from “the HPLC looked clean.”

7. Who ran the test?

Treat the lab name as part of the result:

• Independent, accredited labs (ISO/IEC 17025 is the usual competence standard for testing labs) are stronger than an unsigned in-house printout.
• The report should name the method, instrument, analyst or approver, and sample ID.
• Third-party testing still only covers the sample that was sent. It does not audit the rest of the batch, the warehouse, or the next lot.

In-house COAs are not automatically false. They are easier to template. Look for chromatograms, raw masses, and a lot number you can reconcile.

8. Red flags

Walk away from the document — not toward a different vendor list — if you see:

• No lot number, or a lot that doesn’t match the container
• Purity with no method, no date, and no chromatogram
• Identity by color or “typical HPLC”
• A COA reused across products or dates
• Claims that a research-use COA means “pharmaceutical grade,” “safe for human use,” or “FDA tested”
• Endotoxin or sterility results with no method, no units, and no lab

Bottom line

Read a COA as you would read a study: what was measured, in what sample, with what method, and what was not measured. High HPLC purity is a chemistry result. It is not clinical evidence, not sterile-manufacturing evidence, and not legal permission to use a compound as a drug.