How to Tell If a Peptide Supplier Is Legitimate: The 2026 Verification Checklist

A certificate crossed our desk in August 2026 reporting a purity of 99.998%. The issuing laboratory's own copy of that same report, pulled directly from the laboratory's own domain, said 99.798%. One character had been changed after issue. Everything else on the document was genuine: the letterhead, the task number, the sample name, the analysis date, the chromatogram.

That is the state of peptide supplier verification in a single example. The document was real. The laboratory was real. The number was not.

Most buying guides stop at "make sure there is a certificate of analysis." That advice was useful in about 2019. Today a certificate is the easiest thing in this market to produce, and the gap between suppliers is no longer whether a document exists. It is whether the document can be traced back to the laboratory that supposedly issued it, and whether the numbers on it mean what the reader assumes they mean.

This guide is the checklist we actually apply. It is ordered by how hard each check is to fake, cheapest fakes first, so you can stop early on a bad supplier and spend your time on the ones that survive.

Why "third-party tested" stopped being a filter

Walk through ten peptide storefronts and nine will carry the phrase "third-party tested." It costs nothing to type. It names no laboratory, commits to no method, and refers to no specific batch, which means it cannot be contradicted by any evidence and therefore carries no information.

The phrases that do carry information are the ones that create a checkable fact:

The pattern: a claim is worth reading in proportion to how easily it could be proven wrong. A supplier who publishes a report number has handed you a way to catch them. A supplier who writes "third-party tested" has not.

The nine checks, ranked by how hard each is to fake

Work down this table. Most failing suppliers are eliminated in the first three rows, and those three cost you about five minutes.

Rows 4 and 6 are the load-bearing ones. They are the two a dishonest supplier cannot satisfy cheaply, which is exactly why they are the two most often missing.

Checks 1 to 3: the certificate itself

1. Does a certificate exist at all

"COA available on request" is a soft no. Request it. A supplier who intends to provide one sends it within a day; a supplier who does not will discover a reason. Note which happened, because the answer to this question is usually the answer to every later one.

2. Is it batch-specific

A single certificate covering an entire catalogue is not a certificate. It is a brochure with a chromatogram on it.

What you want: the lot number printed on the vial in your hand appears on the document. If the vial carries no lot number at all, the material has no traceability by construction, and nothing downstream can fix that.

Worth drawing a distinction most vendor pages blur, ours included: a product identifier and a lot number are not the same thing. An identifier derived from the compound and the vial strength is the same string on every vial of that product, so it tells you what the material is meant to be. Only a lot number changes between production runs, and only a lot number ties a vial to one analysis. When you compare a vial against a certificate, the field that has to match is the second one.

3. Is the laboratory named

An unnamed laboratory cannot be checked, which returns you to the "third-party tested" problem one level down. Named laboratories with public registers are the ones worth seeing, because naming one is what makes check #4 possible.

How to prove a certificate is genuine, not merely readable

This is the section that separates this checklist from the rest of them, and it is where we have done the most first-hand work: we built and shipped an automated certificate verifier, and most of what follows was learned by getting it wrong first.

The register lookup

Some laboratories publish a register. The strongest form pairs two identifiers: a task or report number, and a verification key. The lookup returns a record only when both agree. A wrong key against a real task number returns an empty record, not a plausible one, which is what makes the pairing meaningful rather than decorative.

The practical instruction:

1. Find the report number and the verification key on the document.
2. Go to the laboratory's verification URL directly. Type the domain yourself. Do not follow a link the seller provided, because a link is a thing a seller controls.
3. Confirm that the record returned names the same sample, the same date, and the same purity figure as the document you were sent.
4. Where the register serves the laboratory's own image of the report, compare it against your copy field by field.

Step 4 is the one that catches the 99.798 to 99.998 edit described at the top of this page. Steps 1 to 3 would all have passed on that document, because every identifier on it was authentic. Only the side-by-side against the laboratory's copy exposed it.

The provenance rule

The single principle worth carrying away:

A file obtained from the laboratory's own domain is authoritative. A file obtained from the seller is a claim about a file. They are not the same category of evidence, and no amount of internal consistency turns the second into the first.

This is why "they emailed me the PDF and it looks perfect" is weak evidence. A perfect-looking PDF is the expected output of a competent forgery. The check is not whether the document looks right; it is whether it came from somewhere the seller does not control.

What building the verifier taught us

Three findings from that work that apply to a human reading a certificate:

• "We could not check it" is not "we checked and it failed." A register that times out, a blocked request, or a document with no readable text layer are all unknowns. Suppliers get wrongly condemned by readers who collapse the two. Keep them separate in your own notes.
• Do not convict on a misread character. Our own transcription pass once read a verification key as `6C38CRAGL1BC` when the document said `6C38CRAGLTBC`, a single glyph, and the register duly returned "no such test." That one character nearly branded a genuine laboratory report a forgery. If a lookup fails, re-read the key by eye at high magnification before drawing any conclusion.
• A long tail of small oddities is not proof of anything. Real reports from real laboratories accumulate harmless irregularities: odd margins, an abbreviation, a compressed logo. Weight the decisive checks, and do not let eight cosmetic observations outvote one confirmed register match.

Volta publishes the verifier itself as a free certificate of analysis verification tool, and a faster COA red flag checker for a first pass. Both work on any supplier's document, not only ours.

Checks 4 to 5: what the methods actually measure

This is where technically careful buyers still get caught, because the numbers are real and the interpretation is wrong.

HPLC purity, and the three things it cannot tell you

Reverse-phase HPLC with UV detection, usually at 214 nm where the peptide bond absorbs, is the standard purity method. The reported figure is an area percent: the target peak's share of total detected peak area.

Three consequences that the number's presentation actively hides:

1. It only sees what absorbs. Water, inorganic salts, residual scavengers and the counter-ion left over from purification are largely invisible at 214 nm. They are in the vial. They are not in the percentage. This is the single most misunderstood figure in the market.
2. It is relative, not absolute. It reports the composition of the material that reached the detector. It says nothing whatsoever about how much material is in the vial. Purity and quantity are independent, and a certificate reporting only purity has answered only half the question.
3. A fast gradient flatters the result. Closely related impurities, in particular deletion sequences missing a single residue, co-elute under the main peak on a short gradient and resolve on a shallow one. The same sample can report meaningfully different purity under two legitimate methods. A certificate that states its gradient and run time is being straight with you; one that reports a bare percentage is not necessarily lying, but it is not showing its work.

Our HPLC purity interpreter walks a chromatogram field by field if you want to practise on one.

Mass spectrometry confirms mass, not sequence

MS answers one question well: does the observed molecular weight match the expected molecular weight for this compound? A match is meaningful. It is also narrower than most readers assume.

Molecular weight does not determine sequence. Reorder the amino acids and the mass is unchanged to the last decimal. Distinguishing a correct sequence from a scrambled one with identical composition requires MS/MS fragmentation, which is a different and more expensive experiment, and which almost no research-grade certificate includes.

So: MS confirms you have a molecule of the right mass. Combined with HPLC it confirms that molecule dominates what the detector saw. Neither speaks to how much of it is in the vial, which brings us to the check that almost nobody runs.

Check 6: net peptide content, the number almost nobody checks

A lyophilized peptide vial does not contain only peptide. It contains, typically:

• The peptide itself.
• Counter-ions left from purification, commonly trifluoroacetate or acetate, bound to the peptide's basic residues.
• Residual water, measured by Karl Fischer titration.
• Trace salts and scavengers.

Net peptide content is the fraction of the vial's mass that is actually peptide, determined by amino acid analysis or nitrogen analysis. For a typical trifluoroacetate salt it commonly lands between 70% and 85%. It is reported on a serious certificate and absent from a thin one.

Here is what that does to a vial labelled 10 mg, at a stated HPLC purity of 99.2%:

Both figures on that certificate are true simultaneously. The material is 99.2% pure and 82% peptide. Purity describes the composition of the peptide fraction. Net content describes how large that fraction is. A supplier reporting only the first has told you the flattering half.

Two practical notes. Counter-ion identity varies by compound and by manufacturer, so do not assume trifluoroacetate. And where net peptide content is genuinely unreported, treat the vial as an unknown quantity of peptide, not as the labelled quantity, when you plan any quantitative in-vitro work. Our net peptide content calculator does the conversion.

Check 7: price per milligram of actual peptide

The cheapest vial is frequently the most expensive material, and the arithmetic that shows this takes thirty seconds.

Compare three hypothetical offers on the same compound:

Vendor A wins every column a storefront displays and cannot be evaluated on the only column that matters, because the input does not exist. That is not a pricing advantage. It is missing information, and the honest way to record it is as a blank, not as a win.

The general rules:

• Convert everything to price per milligram before comparing anything. Vial sizes differ specifically to make this hard. Our cost per milligram calculator handles a vendor table at once.
• A real price floor exists. Solid-phase synthesis, preparative HPLC purification, lyophilisation and independent analysis have a combined cost. A price far under the market is not a discount on the same product; it is a different product, or a smaller quantity of it, or material that skipped a step.
• Shipping and customs belong in the comparison. A cheap vial with a brokerage fee attached is not cheap.

Checks 8 to 9: who they are, and how it arrives

8. Can you identify the business

A supplier who will not state a legal entity and a physical address has removed your route to recourse before you have paid. This is not a documentation nicety. If the material is wrong, an anonymous storefront is where the conversation ends.

Look for: a named legal entity, a street address that resolves to a real place, a contact route that a human answers, and published terms covering what happens when a shipment is wrong. Two of those are cheap to fake and two are not, which is why you want all four.

9. Does the handling make physical sense

Peptides degrade. The claims worth checking:

• Lyophilized on arrival. Dry powder is markedly more stable than material in solution. A supplier shipping reconstituted material has taken on a cold-chain obligation most cannot meet.
• Stated storage conditions, specific by compound rather than one line for the catalogue. Long-term storage is generally cold and dark; the specifics vary, and our peptide storage guide breaks them down by compound class.
• Customs handled explicitly for international shipments, with the declaration prepared by the sender. Vague answers here are how parcels are lost.
• Protection from repeated freeze and thaw, which is a real degradation pathway once the vial is in your hands rather than the supplier's.

The red flags, and what each one actually indicates

The "investigate, do not convict" row is deliberate. A failed register lookup is the most decisive-feeling result in this entire process and it is the one we have personally seen produce a false accusation. Re-read the key first.

Score a supplier in ten minutes

Two points where the answer is documented and checkable, one where it is asserted but plausible, zero where it is absent.

13 to 16: documentation is genuinely strong. 8 to 12: workable, provided you know which rows are the gaps. Below 8: you are trusting marketing copy. Note that the two bolded rows are worth more than their two points suggest, because they are the two that cannot be satisfied cheaply.

Where Volta scores on its own checklist

Publishing a standard and then exempting yourself from it is the oldest move in vendor content. So, scored honestly, including the row we currently fail:

Total: 8 of 16, which is the bottom of the middle band. Workable, with the gaps known, and the same verdict this page would give any other supplier at that score.

The zeros and ones are real and worth stating plainly:

Our public certificate archive is empty right now. We retired the previous set of published certificates in August 2026 rather than leave superseded results on the site, and the replacements are being re-issued. Until that archive is back, our certificates page shows an empty state instead of documents, and check #1 above is one we fail rather than pass. A batch certificate ships with every order in the meantime, and we will send the document for a specific lot before you buy if you ask at team@voltapeptides.com.

>

Net peptide content is not currently on our published specifications. By the standard set out above, that is the flattering half of the data, and it applies to us as much as to anyone.

>

The batch number on our product pages is a product identifier, not a lot number. It is derived from the compound and the vial strength, so every vial of a given product shows the same string. It is useful for confirming you are looking at the right material and useless for tying a vial to one analysis. The lot number that does that is on the certificate issued with the order.

We would rather you scored us with the gaps visible than trusted a page that scored itself 16 of 16. A supplier's willingness to name its own weakest row is itself a datapoint, and it is one of the few that cannot be manufactured.

Where we are genuinely strong is narrower than a marketing page would claim: the product identifier is derived from the catalogue rather than typed into each record, so a product page and its stated strength cannot silently drift apart, handling is specified per compound rather than once for the catalogue, and our quality process is published in full.

Legality and framing

Research peptides sold in this market are not approved drugs. They are supplied explicitly as laboratory reagents for research use only, and are not for human or veterinary use. Purchasing within that framework is lawful in most jurisdictions; the framework is the condition, not a formality.

Two consequences worth being explicit about:

• A supplier making health or outcome claims is telling you something about the supplier. A vendor who describes benefits, protocols or human quantities has stepped outside the research-use framework, and a vendor careless about that boundary is not usually careful about analytical documentation either. This is the rare red flag that is also a compliance signal.
• Approved pharmaceutical versions of several of these compounds exist and are prescription-only. They are a different product under a different regime, and a research reagent is not a substitute for one.

Related reading

• How to read a certificate of analysis, field by field
• 10 best Canadian peptide vendors in 2026, this checklist applied to a real market
• Free COA verification tool
• Peptide label decoder

Frequently Asked Questions

What is the single most useful check if I only have one minute?

Take the report number and verification key off the certificate and look them up on the laboratory's own domain, typing the domain yourself rather than following the seller's link. It is the only check in this guide that a dishonest supplier cannot satisfy cheaply, and it takes under a minute.

A vendor says "COA available on request." Is that acceptable?

It is acceptable only if you make the request and they answer it. Treat the response time and the specificity of what arrives as the actual test. A batch-specific document within a day is a pass; a catalogue-wide PDF, a delay, or a reason is the answer to every later question too.

Can a certificate of analysis be faked?

Yes, and cheaply. The letterhead, the chromatogram image and the numbers are all editable, which is why a certificate is evidence only when it is traceable to the issuing laboratory. The most instructive forgeries are not invented documents but genuine ones with a single figure altered after issue, which only a comparison against the laboratory's own copy detects.

Why does a vial with 99% purity contain only about 82% peptide?

Because they measure different things. HPLC purity is the target peak's share of the UV-absorbing material the detector sees at 214 nm. Water, salts and the counter-ion bound to the peptide do not absorb meaningfully at that wavelength, so they are present in the vial and absent from the percentage. Net peptide content is the figure that captures them, and it is reported separately when it is reported at all.

Does mass spectrometry prove the peptide is correct?

It proves the molecular weight matches expectation, which is meaningful but narrower than it sounds. Identical amino acids in a different order produce an identical mass, so MS alone cannot distinguish a correct sequence from a scrambled one of the same composition. Sequence confirmation requires MS/MS fragmentation, which most research-grade certificates do not include.

How do I compare prices between vendors fairly?

Convert to price per milligram of actual peptide, not price per vial. Divide the price by the labelled milligrams, then multiply the labelled milligrams by net peptide content where it is reported. Where it is not reported, record the result as unknown rather than as a win, because a vendor who omits the figure cannot be compared on it.

Is a failed register lookup proof that a certificate is fake?

No, and treating it as proof is the most common way an honest supplier gets wrongly accused. A single misread character in a verification key returns exactly the same empty result as a fabricated report. Re-read the identifiers by eye at magnification, and keep "could not verify" separate from "verified as false" in your own record.

Is it legal to buy research peptides?

In most jurisdictions, yes, as laboratory reagents supplied for research use only. They are not approved for human or veterinary use, and the research-use framing is the condition of that legality rather than a disclaimer bolted on afterwards. Approved pharmaceutical versions of some of these compounds exist separately and are prescription-only.

Research use only

This content is provided for informational purposes and does not constitute medical advice. All materials referenced are supplied for research use only and are not for human or veterinary consumption.