How to Read a Peptide Certificate of Analysis (2026)
Share
How to Read a Peptide Certificate of Analysis
By the ThePeptide.ca Research Team · Worked examples use our own published lots · Last updated July 31, 2026
A peptide Certificate of Analysis is a batch-specific laboratory report. Four fields decide whether it can be verified at all: the lot number, the assay method, the measured result, and the named testing laboratory. If any one of those is missing, the document cannot be tied to the vial in your hand and the number on it means nothing. A COA that can be verified independently — by checking the lot against the laboratory’s own records — is worth more than any purity figure printed on a page.
Key takeaways
- Four fields or it’s not evidence: lot number, assay method, measured result, named lab.
- “Purity” answers one question, not three. Identity, net content and net purity are separate tests. Most marketing collapses them into a single number.
- A purity figure is silent on endotoxin, sterility and heavy metals. Those are different tests, and most COAs omit them.
- Certificates can be self-made, borrowed, fabricated or stale. The borrowed kind is the hardest to spot — it verifies as genuine because it is genuine, just not for your vendor.
- A rounded claim is a marketing number. “≥99%” across an entire catalogue is a design element. Real assay results vary.
- Independent verification beats any document. If the trail never leaves the vendor’s own website, nothing has been verified.
Almost every guide to reading a peptide COA is written by someone selling peptides — which is a little like asking a restaurant to write your food-safety inspection. This one is too. So rather than ask you to trust it, it shows you the checks, names the tests, and ends with our own certificates as the worked examples, including the one that came back below 99%.
The four fields that decide whether a COA means anything
A Certificate of Analysis is a report on one specific batch of material. Everything else on the page is presentation. These four fields are what make it checkable:
| Field | What it does | If it’s missing |
|---|---|---|
| Lot / batch number | Ties the document to one specific production run — e.g. RETA-0626-01. |
The report cannot be matched to the vial you received. It is a report about something, somewhere. |
| Assay method | States how the number was produced — e.g. HPLC-UV at 214 nm, or LC-MS for identity. | You don’t know which question was answered. A purity figure with no method is a number without units. |
| Measured result | The actual figure the instrument returned, to the decimal — e.g. 99.68%, 11.88 mg. | A rounded or thresholded claim (“≥99%”) is a marketing statement, not a measurement. |
| Named laboratory | Identifies who ran the test, so the result can be traced back to them. | “Third-party tested” with no lab named is not third-party testing. There is nobody to ask. |
How peptide purity is actually measured
The standard method is reverse-phase high-performance liquid chromatography with ultraviolet detection — HPLC-UV. The sample is dissolved and pushed through a column that separates its components by how strongly each one sticks to the packing material. Compounds emerge at different times and pass a UV detector, which records a peak for each.
Why 214 nanometres? Because that is where the peptide bond itself absorbs ultraviolet light. Detecting at 214 nm means you are seeing essentially every peptide-containing species in the sample, not just the ones with aromatic side chains (which is what you would mostly get at 280 nm). It is the wavelength that gives the most complete picture of what is in the vial.
The purity figure is an area calculation. Software integrates the area under every peak in the chromatogram and reports the target peak as a percentage of the total. So “99.68% purity” means: of everything the detector saw, 99.68% of the absorbance belonged to the peak identified as the target compound.
Three different tests, collapsed into one number
“Purity” is routinely used to mean three separate things that answer three separate questions. This is the single most consequential misreading in the category, and it is why a product can be “99% pure” and still be the wrong substance.
| Test | Question it answers | Typical method |
|---|---|---|
| Identity | Is this the molecule it claims to be? Confirms molecular weight and structure. | Mass spectrometry (LC-MS / MS) |
| Net content / quantity | How much of it is in the vial? A 10 mg vial should contain 10 mg. | Quantitative HPLC against a reference standard |
| Net purity | Of what is present, how much is the target? The percentage figure. | HPLC-UV, peak area |
A vial can pass purity and fail identity. An independent buy-and-test investigation reported a product sold as semaglutide returning a molecular weight of 2,847 Da against an expected 4,113 Da — a different molecule entirely, which no purity percentage would have revealed. Purity tells you how clean the contents are. Identity tells you what they are.
What a purity number is completely silent on
Even a genuine, verified 99% purity result says nothing about contamination. These are separate assays, and most peptide COAs simply do not include them.
Endotoxin
Endotoxins are fragments of bacterial cell wall. They survive sterilisation, are not removed by filtration, and are measured in endotoxin units (EU) rather than as a percentage — which is why a purity assay cannot see them at all. Published analysis of over 6,000 peptide samples found roughly 15% tested positive for endotoxins, and a peer-reviewed test-purchase study found endotoxin present in every delivered sample it analysed.
Heavy metals and residual solvents
Forensic analysis of seized falsified peptides reported 26% containing arsenic or lead above pharmaceutical thresholds, and 100% containing residual tetrahydrofuran, an industrial solvent used in synthesis. An independent multi-vendor test found lead and mercury in budget-tier product alongside purity figures in the 71–83% range.
So the honest framing is: a purity figure is necessary and insufficient. It is one assay among several, and the ones it does not cover are the ones that speak to contamination rather than composition.
↑ Back to topThe four ways a Certificate of Analysis can be false
Not every bad COA is forged. There are four distinct failure modes, and they are not equally easy to spot.
| Type | What it is | How to catch it |
|---|---|---|
| Self-made | A vendor-designed document with lab-style formatting. No independent laboratory was involved. | No lab named, or a “lab” with no independent web presence. No route back to the laboratory. |
| Borrowed | A real laboratory report — issued to somebody else, often an upstream wholesaler, and reused by a reseller. | It verifies as genuine, because it is. Check the client name on the lab’s own record — if it isn’t your vendor, the report isn’t about your vendor’s material. |
| Fabricated | Forged outright — altered figures, invented reference numbers. | Verification link fails or returns nothing. Internal inconsistencies: one investigation found certificates where the testing date preceded the manufacture date. |
| Stale | A genuine report for a genuine batch — that shipped months or years ago. | Compare the COA date and lot number against the lot you actually received. They should match. |
Verifying a COA in about sixty seconds
The point of verification is to leave the vendor’s website. If every piece of evidence lives on the seller’s own server, the only thing attesting to it is the seller.
- Read the lot number off the vial — not off the product page. The physical label is the thing you are trying to verify.
- Find that exact lot in the vendor’s published results. Not a lot from the same compound. The same lot.
- Follow the verification route to the laboratory. A report number, portal record, or direct link to the lab’s own system. If the only copy of the certificate is a PDF hosted by the vendor, you have not verified anything.
- Check the client name on the lab’s record. This is the step that catches a borrowed certificate. The report should have been issued to the vendor selling to you.
- Match the details across both. Lot number, purity figure, measured content, method and date should agree on the vendor’s page and the lab’s record.
- Check the date against what’s shipping. A current certificate should correspond to current stock.
If a vendor cannot survive that sequence, no amount of purity claimed on the page compensates. Run it on us; run it on anyone.
↑ Back to topRed flags
Treat any of these as unverifiable
- No lot number — the document cannot be tied to your vial.
- No laboratory named — “third-party tested” with nobody to ask.
- No assay method or wavelength — a figure with no stated basis.
- An image only — a flat screenshot with no verification route and nothing machine-readable.
- No date, or a date that doesn’t correspond to current stock.
- The same rounded number across the whole catalogue — “≥99%” on every product is a badge, not a result.
- No endotoxin or heavy-metal testing offered at all, on any lot, ever.
- Purity without identity — no mass-spec confirmation of what the compound actually is.
None of these prove misconduct. They prove that you cannot check — and unverifiable is the only standard that matters when you cannot inspect the material yourself.
↑ Back to topWorked examples — our own published lots
Every lot below was tested by Testides (Toronto), an ISO 17025-aligned laboratory, by HPLC-UV at 214 nm, with identity confirmed by mass spectrometry. Apply the checks above to them.
| Product | Lot | Purity | Measured content |
|---|---|---|---|
| Retatrutide 10 mg | RETA-0626-01 |
99.68% | 11.88 mg |
| Tirzepatide 10 mg | TIRZ10-0626-01 |
99.65% | 10.34 mg |
| BPC-157 10 mg | BPC10-0626-01 |
99.64% | 10.99 mg |
| Wolverine Blend | HEAL-0626-01 |
98.64% combined | — |
Why one of these is below 99%
The Wolverine Blend lot returned 98.64% combined purity. It is published as tested, because blends publish actual combined purity as a matter of policy. A blend contains multiple compounds, which makes a combined figure arithmetically harder to push high than a single-compound vial — and the common industry response is to display a generic “≥99%” badge across the entire catalogue instead.
That is the practical lesson for reading anyone’s certificates: real assay results vary, and a catalogue where every product shows the same round number is telling you something about the badge rather than the batch.
Current certificates are published at our lab results page, each linked to the testing laboratory’s own record.
↑ Back to topFrequently asked questions
What is a peptide Certificate of Analysis?
A batch-specific laboratory report describing the identity, quantity and purity of one particular production run of material. Four fields determine whether it can be verified: the lot number, the assay method, the measured result, and the named testing laboratory. Without all four, the document cannot be tied to the vial you received.
What does 99% purity actually mean?
It means that of everything the detector observed in the sample, 99% of the absorbance belonged to the peak identified as the target compound, as measured by HPLC-UV. It does not confirm the compound’s identity, does not state how many milligrams are in the vial, and says nothing about endotoxin, sterility or heavy metals. Those are separate tests.
Why is peptide purity measured at 214 nm?
Because the peptide bond itself absorbs ultraviolet light at approximately 214 nm. Detecting at that wavelength captures essentially all peptide-containing species in the sample, rather than only those with aromatic side chains, which is what detection at 280 nm would largely show.
Can a peptide COA be faked?
Yes, in four distinct ways. Self-made certificates are vendor documents with no laboratory involved. Borrowed certificates are genuine lab reports issued to a different company and reused. Fabricated certificates are forged outright, sometimes with internal contradictions such as a testing date preceding the manufacture date. Stale certificates are real but describe a batch that shipped long ago.
How do I verify a COA is real?
Read the lot number from the vial, find that exact lot in the vendor’s published results, follow the verification route to the testing laboratory’s own record, check that the client name on that record is the vendor selling to you, and confirm the lot number, purity, method and date match on both. If the certificate exists only as a file hosted by the vendor, nothing has been independently verified.
What is the difference between HPLC and mass spectrometry on a COA?
HPLC separates the components of a sample and, with UV detection, measures how much of the total is the target compound — this produces the purity percentage. Mass spectrometry measures molecular weight and confirms identity, answering whether the substance is the molecule it claims to be. Purity and identity are different questions and require different tests.
Why is endotoxin testing rarely on a peptide COA?
Because it is a separate assay with additional cost. Endotoxins are bacterial cell-wall fragments measured in endotoxin units; they survive sterilisation and are not detected by a purity assay. Published analysis of over 6,000 peptide samples found roughly 15% positive for endotoxins, and a full panel including endotoxin and heavy metals adds several hundred dollars per batch.
Does a high purity number mean the product is safe?
No. Purity describes composition, not contamination. A sample can return a high purity figure while containing endotoxin, heavy metals or residual solvents, none of which a purity assay measures. Forensic analysis of seized falsified peptides reported 26% containing arsenic or lead above pharmaceutical thresholds and 100% containing residual tetrahydrofuran.
Should I trust a vendor’s own COA?
Only to the extent it can be independently verified. The useful question is not whether a vendor can show you one good certificate, but whether the lot on your vial traces to a named laboratory’s own record, whether that record was issued to the vendor rather than an upstream supplier, and whether the vendor publishes results consistently across batches over time rather than once.
Related research resources
- Our published lot COAs — each linked to the testing laboratory’s own record
- Where to get research peptides tested in Canada — running your own independent test
- Buying research peptides in Canada — full supplier evaluation criteria
- Are peptides legal in Canada? — the regulatory position, from primary sources
- Storage and stability — why a COA date matters
Sources
- Nature. Analysis of the research peptide market, including testing of more than 6,000 samples (approximately 40% failing basic purity standards; approximately 15% positive for endotoxins). 2026. nature.com.
- Journal of Medical Internet Research. Multifactor quality and safety analysis of semaglutide products purchased from illegal online pharmacies. 2024. jmir.org.
- Independent multi-vendor buy-and-test investigation reporting purity, endotoxin and heavy-metal results across vendor tiers, including a product returning a molecular weight of 2,847 Da against an expected 4,113 Da. thepeptidelist.com.
- Forensic analysis of seized falsified peptides (purity range, heavy metals above pharmaceutical thresholds, residual tetrahydrofuran). Summarised at glpwinner.com.
- Guidance on identifying self-made, borrowed, fabricated and stale certificates, including verification of the client field on laboratory records. thepeptidewatch.com.
- ThePeptide.ca published lot certificates, tested by Testides (Toronto), HPLC-UV at 214 nm. thepeptide.ca/pages/lab-results.
Changelog. July 31, 2026 — Rebuilt as a dedicated COA guide. Supplier-evaluation content moved to the buyer’s guide. Added: how HPLC-UV at 214 nm produces a purity figure; the identity / net content / net purity distinction; what purity does not measure, with published endotoxin and heavy-metal findings; the four false-certificate types including the borrowed-COA client-name check; a six-step verification sequence; a red-flag list; and worked examples using our own four published lots, including the blend at 98.64%.