How to choose a peptide testing lab: 12 checks before you send a vial

What a peptide test is, and what a lab must prove

A peptide test is a set of lab measurements on one vial. It answers four questions. Which peptide is in it (identity)? How many mg of it are there (content)? What share of the UV-absorbing material is the target peptide (purity)? And, if you order those tests, does it carry contamination such as endotoxin or live microbes? A good lab proves each answer on the report itself, with the method, the unit and the raw data.

So the way to choose a peptide testing lab, or to compare peptide testing services, is to read their reports before you send a vial. Ask the lab for a sample report, or open one a seller has published, and run it through the 12 checks below. They come from three places: our review of public certificates of analysis (COAs) from 26 report formats (25 peptide testing labs plus Finnrick’s certificate format) in September 2026; ISO/IEC 17025 clause 7.8, the reporting rules for test labs; and the criteria that rating sites publish.

The 12 checks at a glance

#	Check	What to look for on a sample report	Seen on the 26 report formats we read
1	Identity proven by mass	Expected mass, observed mass and the error in ppm	12 show a spectrum; 3 print both masses with an error
2	Purity and content as two numbers	% peak area and mg per vial, with the basis of the mg	21 report mg; 1 states the basis clearly
3	Raw data	Full chromatogram and a peak table with areas	4 print a proper peak table
4	Endotoxin as a number	EU per mg or per vial, limit of quantitation, spike recovery	None in full; 11 in part
5	Sterility that is USP <71>	At least 14 days of incubation, units tested	None documented in full
6	Verification at the lab	Report ID plus a key that cannot be guessed	12 link to a lab copy; 6 use unguessable codes
7	Lab-taken vial photo	Photo taken on receipt, not supplied by the client	16
8	Client relationship stated	Who ordered, and any link to the lab	None
9	Corrections as new versions	Version number; old versions marked	1 in full, 1 in part
10	Measurement uncertainty	A ± value with its coverage factor (k)	1 in full, 3 in part
11	Published turnaround	A figure, and when the clock starts	A website fact, not a report field
12	Accreditation with scope	Certificate number and a scope that covers the test	20 made no over-claim

The right column counts what the reports show. It does not measure whether a lab’s numbers are right: a lab can print every field and still measure badly. Use the checks to see what a report proves. The same checks, lab by lab and with sources, are in peptide testing labs compared.

Checks 1 to 3: the core numbers

Check 1: identity proven by mass

Retention time alone is not specific. ICH Q6A, the international guideline on specifications, says so, and GLP-1 peptides elute close together on a column. Mass spectrometry measures the mass of the molecule itself. Semaglutide, tirzepatide and retatrutide differ by about 80 Da or more, so a mass match tells them apart. A good report prints the expected mass, the observed mass and the error in ppm, so you can check the match with a calculator. A spectrum picture with no expected value is weaker evidence. How it works: LC-MS identity.

Check 2: purity and content as two numbers

Purity is the main peak’s share of the UV peak area. Content is the mg of peptide in the vial, measured against a reference standard of known content. They answer different questions: a vial can be 99% pure and hold 6 mg where the label says 10.

The report should also state the basis of the mg. Net peptide (free base) leaves out the counter-ion salt and water. A result on total powder reads higher. Peptigrity, which collects COAs, never shows a result above 110% of the label as good, because gross-powder reporting inflates it. About 40% of the certificates it held were above 110% in September 2026. In our review, 21 of the 26 formats reported mg per vial, and only one stated plainly what that mg meant.

Check 3: the raw data behind each number

A purity figure without its chromatogram is a claim. Look for the full chromatogram (not cropped to the main peak), a peak table with areas, and the mass spectrum behind the identity result. Only 4 of the 26 formats had a proper peak table, and none printed its integration settings. Some labs sell raw data as an add-on, so ask what the standard report includes. Our example report shows what a full set looks like, and how to read a COA explains each part.

Checks 4 and 5: safety tests named for what they are

Check 4: endotoxin as a number

Endotoxin is measured with the LAL test and reported in endotoxin units (EU). A useful result gives EU per mg and EU per vial, the limit of quantitation, and the spike-recovery control that shows the sample did not mask the result. “Pass” alone hides the number. A research peptide has no labelled dose, so no official limit applies to it. A report that prints a limit should say whose limit it is. No report we read met all of this, and 11 labs met part of it. More: endotoxin testing for peptides.

Check 5: sterility that is the USP <71> test

Sterility under USP <71> means incubating the sample in two culture media for at least 14 days and checking for growth. Labs also sell microbial counts (TAMC and TYMC), PCR screens for microbial DNA and short cultures under the name “sterility”. These answer other questions. None of the reports we read documented a 14-day USP <71> test with the number of units tested, although several labs list it as an option. The result covers only the units tested, so ask how many. Detail: sterility testing.

Checks 6 to 9: can you trust the document

Check 6: verification at the lab, with a key no one can guess

Open the report on the lab’s own site, not from a seller’s PDF. Good systems pair a report ID with a random access key or QR code and return the same file the lab issued. Weaker ones use sequential numbers that anyone can step through, or an upload form the lab answers by email. In our review, 12 labs linked to a lab-hosted copy, and 6 used codes that cannot be guessed. Rating sites reward this. Disclosed Labs adds 10 points for a public verification portal, and Peptigrity marks each certificate as lab-verified or not. Steps: how to verify a peptide COA.

Check 7: a photo of your vial, taken by the lab

A photo taken on receipt lets you compare cap colour, fill level and printed lot with your own vial. It must be the lab’s photo, not one the client sent in. 16 of the 26 formats showed one.

Check 8: the client relationship, stated

Who ordered the test, and do they have any link to the lab? None of the 26 formats stated this. Rating sites treat independence as a gate. Merit Verified says vendor-owned labs are flagged and cannot earn a passing tier, and Peptigrity does not accept a vendor’s in-house testing. Detail: independent peptide testing.

Check 9: corrections as new versions

Every lab makes mistakes. What matters is how it fixes them. A good report carries a version number, and a correction is a new version that names the old one and says why. One lab in our review showed a public change history. At several others, files were replaced at the same link with no visible version.

Checks 10 to 12: limits, timing and credentials

Check 10: measurement uncertainty

Every measurement has an error. A report should print it as an expanded uncertainty with its coverage factor (for example “± 0.4 mg, k = 2”) and round the result to match. Only one lab in our review did this in full. Finnrick publishes the answers its partner labs gave to a 10-question method questionnaire. The margins of error stated there run from about 2–3% to 10%, and one lab says it has not established one. A purity printed to three decimals with no uncertainty claims more precision than the method can support.

Check 11: published turnaround, and when the clock starts

Published times for identity, purity and content run from an average of 1–3 business days to 9–11 business days at the labs we checked. Check whether the clock starts on receipt, on log-in or on payment, and whether anyone shows measured times. Peptigrity lists each lab’s stated turnaround next to the turnaround buyers report. Full table: how long peptide testing takes.

Check 12: accreditation, with your test in its scope

ISO/IEC 17025 accreditation covers only the tests listed in a lab’s scope. A certificate can cover food microbiology or solvents and leave out peptide purity, identity, endotoxin and sterility. If a lab claims accreditation, ask for the certificate number and the scope document, find your test on it, and look the number up in the accreditation body’s directory. Merit Verified gives accreditation the most weight in its lab score. A lab without accreditation should say so plainly.

Two questions a report cannot answer

Will your seller accept the result? Many sellers honour refund claims only with results from labs they name. If you test with one lab and the seller accepts only another, you may pay twice. Ask before you order.

What does the price include? Raw data, endotoxin, sterility and extra vials are often priced separately. Published prices are in peptide testing cost.

How rating sites judge labs

Site	What it rewards or penalises
Finnrick	Publishes each partner lab’s test count, pass rate on recent tests and questionnaire answers; says it estimates lab bias from overlapping tests but does not publish the offsets
Disclosed Labs	Starts each lab at 50 points; adds points for COA volume, vendor diversity and a public verification portal; takes 25 off for a very narrow purity spread from few vendors
Merit Verified	Accreditation, HPLC plus LC-MS methods, independence (vendor-owned labs cannot pass), documented integrity concerns
Peptigrity	A verification portal, stated against reported turnaround, a stated mg basis; vendor in-house testing not accepted
PeptideBenchmark	Share of COAs with a working lab link, reused records, whether lookup links still respond

A pass rate on a rating site describes the samples a lab was sent as much as the lab itself. Read each site’s method page before you treat its score as a verdict on a lab. The sites side by side: peptide rating sites compared.

What HK Analytical does on these checks

Our identity test is LC-MS, and the report prints the expected and observed mass. Purity is by HPLC, content per vial is a separate result, and every number comes with its raw data: chromatograms and spectra. Endotoxin (LAL) is reported in EU per mg and per vial, and our sterility test is Sterility (USP <71>, 14 days). Each report is a signed record (an Ed25519 signature over the canonical record), carries our own photo of the vial as received, and states how the client is related to HK Analytical inside the signed record. Each report has a private link and can be checked at hkanalytical.com/verify. Reports are never edited: a change makes a new version, and the old one is marked superseded. We do not hold ISO/IEC 17025 accreditation. Results within days of receipt. Sterility (USP <71>, 14 days) is the one test that takes 14 days, because the method incubates for 14 days. See how we compare and all tests.