Key Takeaways
- •Just under half of Canadian storefronts and seven in ten US storefronts showed a certificate somewhere we looked. That is 14 of 30 in Canada and 32 of 46 in the United States, counting dedicated certificates pages and sampled product pages together.
- •A "certificates page" count alone undercounts. Dedicated pages accounted for 10 Canadian and 23 US stores. Another 4 and 9 kept certificates only on product pages, which a page-address check never sees.
- •Naming a laboratory is common in Canada and patchy in the US. In Canada, 13 of the 14 stores showing a certificate also named a laboratory in readable text. In the US it was 15 of 32.
- •Few stores give a direct route to the laboratory's own record. Only 9 of 76 linked to a specific report on a laboratory's website: 5 in Canada and 4 in the US.
- •Eleven laboratories were named, and one dominates. Janoshik appeared on 14 stores across both countries. No other laboratory appeared on more than 5.
- •None of this measures what is in a vial, and none of it shows that any store skips testing. It measures what a reader can see before ordering. Volta wrote the checks, would pass them, and publishes batched 3rd party lab certificates across its products.
Summary
We took the 76 storefronts from our October 2026 price collection, 30 assigned to Canada and 46 to the United States, and asked four questions a reader can answer without buying anything. Is there a certificate? Where is it? Is a laboratory named? Can you reach the laboratory's own record of the report?
We tried 13 common certificate addresses on every site, read every page that answered, and then read up to three product pages per store, 191 in all. We found a certificate on 14 Canadian and 32 US storefronts, a named laboratory in page text on 14 and 15, and a link to a specific report on the laboratory's site on 5 and 4. These are floors. Certificates can sit behind an account, on product pages we did not sample, or in an inbox, and a laboratory's name inside an image is invisible to a text search. The useful step for any reader is the same whatever the store: find the report, match it to the lot, and confirm it with the laboratory.
Almost every peptide storefront says its peptides are third-party tested. Fewer show you the result, and fewer still let you check that result with the laboratory that produced it. We wanted numbers rather than impressions, so we checked 76 storefronts in Canada and the United States.
This is a study of disclosure, not of quality. We did not buy or test a single vial. We counted what a reader can find on a store's own website before ordering: a certificate of analysis, the name of the laboratory, a link to that laboratory's own copy of the report, and a lot or batch code. A store that publishes nothing may test every lot. A store with a large certificates page may still leave the most important question unanswered. We report what is visible and stop there.
Why Count Disclosure at All
Most published work on peptide quality measures the contents of a vial. That work matters, and it is worth knowing what it found.
- Illegal online pharmacies. A 2024 test-purchase study of GLP-1 sold without a prescription bought from six illegal online pharmacies and received three vials. Measured purity was 7.7% to 14.37% against 99% on the label, content ran 28.56% to 38.69% above the labelled amount, and endotoxin was detected in every sample (Ashraf et al., Journal of Medical Internet Research, 2024). These were pharmacy-style sellers, not research storefronts, but a confident label beside a different measurement is why documentation matters.
- Falsified peptide products in Belgium. The country's official medicines control laboratory built an LC-MS/MS screen for 25 peptides seen by regulators and customs (Vanhee et al., Talanta, 2015). A later systematic screen of the ten most frequently encountered falsified peptide drugs, bought from three suspected illegal internet pharmacies, reported purity of 5% to 75% for cysteine-containing peptides and found arsenic and lead in some samples (Janvier et al., Talanta, 2018).
- A large public testing dataset. A 2026 preprint analysed 6,441 samples of 14 compounds from the Finnrick public dataset. Depending on which of two quality frameworks was applied, 41.6% to 71.1% of samples failed basic quality criteria, and measurable endotoxin was present in 15% of the samples tested for it. The authors note that consumer-directed testing "captures only a small fraction" of what matters (Mendias and Awan, Preprints.org, 2026). It is not peer reviewed. A third-party re-analysis reproduced the two headline rates at 42.2% and 70.8%, then split them into separate measures, for example wrong identity in 0.69% of records, which shows how much a single "failure" figure depends on its definition (TitrateLab, 2026).
All of this shares one property: someone chose which vials were tested. A vendor picks what to submit, a buyer picks what to send, a regulator analyses what was seized. Finnrick, whose methodology page reported 12,729 tests when we read it on 2 November 2026, scores whether a batch identifier appears on the container, on an insert or only on a website, because that link between paper and vial is where checking starts (Finnrick methodology).
Meanwhile buyers lean more on the paper. An on-chain analysis of payments in this market estimated that testing spend per buyer fell by about 88%, and that recent buyers rely mostly on vendor-posted certificates, which typically report mass and purity and may not cover sterility (Chainalysis, June 2026). If the certificate a store posts is now the main evidence most readers see, it is worth counting how many stores post one, where, and how checkable it is. We found no published count of that, so we made one.
The Sample and the Method
Where the 76 came from. Our October 2026 price collection started from 153 domains, read 95 storefronts and matched 3,994 comparable listings. Each storefront was assigned to a market from evidence on its own site, such as its currency, a published address and the carrier named in its shipping copy, never from its domain ending. That gave 30 Canadian and 46 US storefronts. No price from that collection appears here. The sample is the storefronts we could read a catalogue from, which is not every store in either country.
Pass one, 1 November: certificates pages. A script requested 13 common addresses on every site, among them /coa, /coas, /lab-results, /certificate-of-analysis, /third-party-testing and their /pages/ variants. A page counted if it loaded and used words such as "certificate of analysis" or "lab result", and the script also searched it and the homepage for laboratory names. We then read every page it counted: the final address after redirects, the title, and how many PDF links and certificate images it carried.
Pass two, 2 November: product pages and verification. For each store we took up to three product addresses from the published price data, spread across its catalogue, and read them along with every confirmed certificates page. On each page we recorded laboratory names in the text, links to a laboratory's own website, and lot or batch codes written as text. Where a store had no confirmed certificates page but a product page carried a document, we checked the file names and image sources by hand, and opened three ambiguous PDFs. Two were laboratory certificates; the third, a scanned file, could not be read and was not counted.
What counts as what.
- Certificate found: a confirmed certificates page, or a certificate image, PDF or laboratory report link on a sampled product page.
- Laboratory named: an actual laboratory's name in readable page text. Accreditation bodies such as A2LA and generic phrases such as "analytical lab" do not count.
- Link to the laboratory's record: a link to a specific report on the laboratory's own site, for example a report page or a verification address carrying a code. A link to a laboratory's homepage does not count.
- Lot code in text: a lot or batch code with a number in it, written as text beside a report.
What We Found: How Many Stores Show Third-Party Test Results
| Measure | Canada (30) | United States (46) |
|---|---|---|
| Dedicated certificates page, confirmed by reading | 10 (33%) | 23 (50%) |
| Certificate only on a sampled product page | 4 (13%) | 9 (20%) |
| Certificate found anywhere we looked | 14 (47%) | 32 (70%) |
| Nothing found where we looked | 16 (53%) | 14 (30%) |
| A laboratory named in readable text | 14 (47%) | 15 (33%) |
| Link to a specific report on the laboratory's site | 5 (17%) | 4 (9%) |
| Lot or batch code written as text | 3 (10%) | 5 (11%) |
Two notes on the bottom rows. Of the 16 Canadian and 14 US stores where we found nothing, 3 and 7 returned no readable product page at all, so for those stores the product-page half of the method did not run. And two US domains appear to be one store under two addresses; both are counted, so the US certificate figure may be one too high.
Where certificates live
The address a store uses differs by country. Six of the 10 Canadian certificates pages sat at /lab-results. In the US, 16 of 23 sat at /coa or /coas.
Format differs too. Eight of the 10 Canadian pages presented certificates as images and only one linked PDFs. In the US, 15 of 23 linked PDFs and 13 showed images, often both. The median confirmed page carried 9 document links in Canada and 18 in the US, ranging from none, on pages that load through scripts, to several hundred. Links are not reports: one report often appears as a thumbnail and a full file.
Product pages carried more than we expected. Thirteen stores with no dedicated certificates page showed a certificate beside the product itself: a report image, a PDF named with a lot code, or a link to the laboratory's record. The two PDFs we could read were both laboratory certificates, each printing a sample code or lot number. Any study that only checks for a certificates page will miss this group.
Seven pages that a script would count did not hold up when read. Three sent a reader to an account or registration page, so whatever sits there is not public. Two were articles about certificates. One was a general marketing page, and one came back empty when we looked again.
Which laboratories are named
Counting only names in readable text, across every page we read, eleven laboratories appeared.
| Laboratory | Canadian stores | US stores | Total |
|---|---|---|---|
| Janoshik | 9 | 5 | 14 |
| Testides | 5 | 0 | 5 |
| Chromate | 0 | 5 | 5 |
| Freedom Diagnostics | 1 | 4 | 5 |
| Vanguard | 1 | 3 | 4 |
| Accumark | 0 | 3 | 3 |
| ILS Laboratories | 0 | 2 | 2 |
| MZ Biolabs | 0 | 2 | 2 |
| Krause | 1 | 0 | 1 |
| Peptide Test Canada | 1 | 0 | 1 |
| VialTests | 1 | 0 | 1 |
Some stores named more than one laboratory, so the columns add up to more than the store counts. Nine of the eleven have an entry in our testing labs directory, which records what each one offers and how its reports can be checked. We do not rate laboratories here.
The country split is the clearest pattern in the study. In Canada, a store showing a certificate almost always named the laboratory in text: 13 of 14. In the US the figure was 15 of 32. That does not make US certificates anonymous: the two PDFs we read from US product pages each carried a laboratory's letterhead. The name is inside the document rather than on the page.
Can a reader check the report?
A certificate is a claim by whoever posts it. The strongest thing a store can add is a route to the laboratory's own copy, so the reader does not have to trust the store's image of it.
Nine storefronts linked to a specific report on a laboratory's site: report pages at Janoshik (4 stores) and Testides (3), and coded verification addresses at Chromate, Accumark, Peptide Test Canada and VialTests (1 each). A few more linked only to a laboratory's homepage, which says where to go but not which report to find.
Lot codes written as text were rarer still, on 8 stores. That figure is low for a known reason: most lot numbers are printed inside the certificate image, where a text search cannot read them. It is still the number that matters to a reader holding a vial, because the lot on the label and the lot on the report are the two halves of the only check that needs no trust at all.
We did not click through each verification link to confirm the report it returned. That is the natural second edition of this study.
What This Does Not Show
It does not show that any store skips testing. We have no evidence for that and we are not suggesting it. A store can test every lot and publish nothing, or send reports on request.
It does not grade any certificate. We did not check purity figures, methods, dates or whether a report matches the lot on sale. A certificate can be real, current and still describe a different vial from the one shipped.
It undercounts. We tried 13 addresses and three product pages per store. Certificates can sit on other product pages, behind an account or in an email, and a laboratory's name or a lot code inside an image is invisible to a text search. Eight confirmed certificates pages loaded most of their contents through scripts, which hides text from this method.
It is a snapshot of one sample. The 76 stores are the ones whose catalogues we could read in October, not a census, and sites change weekly.
It says nothing about safety panels. Some laboratories in the table above offer endotoxin, sterility and heavy-metal testing, and a few stores mention them. We did not count those results because we did not open the documents.
How to Reproduce This
- Take a list of storefronts and assign each to a market from what its own site says, not its domain.
- Request the 13 addresses listed above on each site. Keep a page only if it loads, uses certificate language, and is still a certificates page after redirects. Read the final address and title by hand.
- Open a few product pages per store and look for certificate images, PDFs and links to a laboratory's site.
- Record laboratory names that are actual laboratories, and separate links to a specific report from links to a homepage.
- Write down what each step cannot see.
Our first script, run in October, counted 12 Canadian and 28 US certificates pages. The November re-run counted 13 and 27, within one store each way, and reading those pages brought them to 10 and 23. The October raw file was not kept, so those two figures cannot be re-derived. We keep the per-store record and both scripts. We do not publish the per-store file, because it would let a reader rank individual stores on a method that cannot support a judgement about any one of them.
What We Would Check Instead
A certificates page is a starting point, not an answer. These questions tell a reader more than any count in this study.
- Does the report name the laboratory, a date, a method and a lot or sample code?
- Does that lot appear on the vial label or the packing slip?
- Does the report show identity as well as purity? HPLC gives a purity figure; a mass spectrometry result shows the molecule is the one named. Our guide to identity testing explains the difference.
- Can you find the same report on the laboratory's own site, from a link or a code, without relying on the store's copy?
- If there is no public report, will the store send one for your lot when asked?
How to read a certificate of analysis explains each field, our supplier checklist covers the wider questions, and peptide testing in Canada covers sending a sample yourself.
Where Volta Stands
Volta would pass every check in this study. We publish a certificates page, name the laboratory, print the lot and link each report to the laboratory's own record. We also provide a batch number and QR code on each vial that connects that specific vial to the batch it came from and the COA for that batch.
We publish certificates publicly. Unlike many vendors we publicly display all of our COAs, we don't request you to contact us to see them. They are all available on our COA page, our quality page and displayed on the product page they correspond to.
All certificates come from a independent 3rd party laboratory. SideChain Analytics ran them by HPLC-MS/MS. Each record carries a lot number and a link to the report on the laboratory's website, so a reader can confirm it there rather than with us.
A report covers one sample. Each certificate describes the sample the laboratory received. Our greater than 99% figure is a release specification, the threshold a lot must meet, not a measurement of every vial. Most vials come in higher than our required minimum of 99%. For example, BPC-157 10mg (batch VPBP10100) came in at 99.9% purity, GLP-3 RT 20mg (batch VPRT20100) 99.7% purity, GHK-CU 100mg (batch VPGC100100) came in at 99.8% purity, GLP-2 TZ 30mg (batch VPTR30100) 99.8% purity. We pride ourselves in having some of the leading purity standards in the industry.
All Volta products are for research use only. They are intended for in-vitro laboratory work and are not for human or animal use.
Frequently Asked Questions
How many peptide vendors publish certificates of analysis?
In our sample of 76, we found a certificate on 14 of 30 Canadian storefronts and 32 of 46 US storefronts, counting dedicated certificates pages and sampled product pages. Dedicated pages alone accounted for 10 and 23. These are floors, because certificates can sit in places we did not look.
Where do stores usually put their lab results?
Canadian stores in our sample most often used a /lab-results page. US stores most often used /coa or /coas. Thirteen stores had no dedicated page and showed certificates on product pages instead, so it is worth checking the product page as well as the menu.
Which laboratories test peptides for stores?
Janoshik, named in readable text on 14 of the 76 stores. Ten other laboratories appeared on between one and five stores each. Naming a laboratory is not the same as linking to its record, and only 9 stores did the second.
How can I verify a peptide certificate of analysis?
Look for the laboratory's name and a report number or code, then find the same report on the laboratory's own website, ideally from a direct link. Compare the lot on the report with the lot on your vial. If the laboratory offers no lookup, write to it with the report number. Our testing labs directory notes how each laboratory handles verification.
How much does it cost to get a peptide third-party tested?
Among the US laboratories in our testing labs directory that publish prices, an HPLC purity analysis of one common peptide costs about US$180 to US$250 per sample. GLP-1 compounds and blends cost more, up to US$400 for one compound and US$800 for a blend. Endotoxin and sterility are separate tests at about US$200 to US$300 each. Finnrick tests fifteen named compounds for free. These are the laboratories' own published prices when we last checked them.
How can I test whether a peptide is real?
Send a sample to an independent laboratory and ask for identity as well as purity. HPLC gives a purity figure but does not prove which molecule is in the vial; mass spectrometry does. Our guide to identity testing explains why, and the testing labs directory lists laboratories that accept samples.
Does a store without public certificates skip testing?
A store without public certificates is a red flag. A store that requires you to contact them to see certificates is also a red flag. If a vendor is proud of their quality standards and is transparent, they would have no reason to hide their COAs.
Does a certificates page mean a store's testing is reliable?
No. It means reports are available to read. A report still has to be matched to your lot (in other words the batch that the vial you received comes from) and confirmed with the laboratory.
What is the difference between this study and Finnrick or a published purity study?
Those measure what is in a vial, from samples someone chose to test. This study measures what a store shows a reader before ordering, across a fixed list of stores. The two answer different questions and neither replaces the other.
Why are no stores named?
The method cannot support a judgement about any one store. A certificate we did not find may sit somewhere we did not look, so naming stores would claim more than we measured. We report laboratories by name because counting them says nothing about any store's practice.
