Twelve red flags on a peptide COA
Most weak certificates are not forgeries. They are real documents that omit the conditions which would let you evaluate them, and omission is much harder to spot than invention.
A certificate is supposed to make a result checkable by someone who was not in the laboratory. That only works if the document carries the things a reader needs to check it: what was measured, on what instrument, under what conditions, against what limit, and with what stated bounds on the conclusion.
Strip those away and you are left with a document that looks authoritative and asks to be trusted. Here are twelve specific absences worth noticing, each with the question to send back.
No detection wavelength anywhere on the document
The same sample gives different area percentages at different wavelengths. A purity figure without the wavelength it was recorded at cannot be compared with any other purity figure, including a later one on the same material.
Ask: What wavelength was this recorded at, and why that one for this analyte?
No column, gradient, or run length stated
These are the method identity. A different gradient on the same column is a different method that can legitimately produce a different number, so a purity figure with no method behind it is not reproducible even in principle.
Ask: Can I have the full method conditions for this run?
A chromatogram with an unlabelled y-axis
Without a scale, small peaks cannot be assessed. A trace can be plotted so the principal peak fills the frame and everything else is compressed into the baseline, which flatters a sample without stating anything false.
Ask: What is the response scale on this plot?
A run that ends immediately after the principal peak
The region after the target is where hydrophobic impurities elute. A method that stops there has chosen not to look, and the resulting purity figure is a statement about a truncated window.
Ask: What was the total run time, and what eluted after the target?
“Pass” with no specification named
A pass is only meaningful against an acceptance criterion and a stated source for it. Without one, “pass” is a claim about nothing; the laboratory has decided on your behalf what threshold matters.
Ask: Pass against what limit, set by whom?
“Not detected” with no reporting limit
Not detected never means none present. It means none present above the method’s reporting limit. Without that number, the statement carries no information about how sensitive the measurement was.
Ask: What is the reporting limit for this analyte on this method?
A purity percentage presented as strength or dose
Area purity is a ratio of detected response and is not milligrams. A certificate that lets a percentage stand in for a content figure is either confused or is inviting the reader to be.
Ask: Was content measured, and if so on what basis and against which reference standard?
A content figure with no calculation basis
Peptide base and gross salt give two different numbers for the same vial. Without the basis, a milligram figure cannot be compared with a label claim or with another laboratory’s result.
Ask: Is this peptide base or gross salt, and what reference standard was it calibrated against?
No sample identifier, lot, or receipt date
A COA is a record about one specific sample. Without identifiers and dates it cannot be tied to anything, which makes it transferable to any vial that looks similar, the exact property a certificate should not have.
Ask: Which specific unit does this document describe, and when was it received?
A report number that does not resolve anywhere
A verification page lets someone who did not receive the original file confirm the document matches the laboratory record. A report number with no registry behind it is a formatting convention, not a control.
Ask: Where can this report number be independently checked?
Conclusions broader than the testing performed
Language about safety, sterility, quality, or suitability appearing on a document that measured identity and purity is scope creep. The testing did not address those things and the certificate should not imply otherwise.
Ask: Which measurement on this report supports that sentence?
The laboratory also sold you the material
An in-house result is not third-party testing. It may be perfectly accurate, but it is not independent, and independence is the specific property most people are trying to buy when they ask for a COA.
Ask: Who performed this analysis, and what is their relationship to the seller?
What none of this establishes
A certificate that clears all twelve is a well-made record of a specific measurement on a specific sample. It is still not evidence of clinical safety, efficacy, sterility, regulatory status, or consistency across untested units, and a certificate that claims otherwise has tripped flag eleven regardless of how good the rest of it is.
The asymmetry worth remembering
Every item above costs a laboratory nothing to include. Detection wavelength, run length, reporting limits, and calculation basis are all already known at the moment the report is written; putting them on the page is a formatting decision, not additional work.
Which means their absence is informative. A laboratory that omits the conditions under which its numbers were produced has made a choice about how much scrutiny it wants its numbers to receive.
Further reference
Sources and related guidance for this article:
- CPT certificate verification
Check the published record and its current status.
- ICH Q2(R2): Validation of Analytical Procedures
Framework for evaluating analytical procedure performance; this reference does not certify CPT or an individual result.
Common questions
How can I tell if a peptide COA is real?
Check that it resolves. A genuine public certificate carries a report number and normally a QR code that opens the issuing laboratory’s own verification record, and the analyte, dates, and report ID on the live page should match the document in your hand. Beyond authenticity, check that the document states its method conditions, reporting limits, and the calculation basis for any quantitative figure. A real but uninformative certificate is a common outcome.
Is a COA from the seller worth anything?
It can be accurate, but it is not third-party testing, and independence is usually the property a buyer is trying to obtain. An in-house or supplier-provided certificate should be read as the seller’s own record rather than as external verification of it.
What should always appear on a peptide certificate?
Sample identifiers and dates, the named method for every result, detection conditions, results with their units and calculation basis, reporting limits, the reviewer or releasing authority, an explicit statement of limitations, and a way to verify the document independently.