Are peptides from China safe? What published lab results show about quality

Are peptides from China safe: what a lab result can answer

No lab test can tell you that a peptide is safe to use. A test measures what is in one vial: which peptide it is, how pure it is, how many mg it holds and, when ordered, how much endotoxin it carries and whether microbes grow from it. It is not a medical judgement, and it says nothing about the vial after it leaves the lab.

Published data can answer a narrower question: do peptides from sellers in China fail those measurements more often than peptides from other sellers? In Finnrick’s published data, they do not. Sellers listed in China failed 26.5% of their classifiable tests, against 30.3% for all sellers. They also hold many of the top places on the product leaderboards. The spread between individual sellers is much wider than the gap between countries.

What 11,999 published tests show by seller location

Finnrick is a US testing and rating platform. By 25 September 2026 it had published 11,999 tests on 34 products from 466 sellers. It sends anonymised samples to partner labs and classes each result as pass, fail or incomplete. Under its default standard, Purity Focus, a sample passes when the identity is right, HPLC purity is at least 99.5% and the fill is within ±35% of the label.

Finnrick’s summaries do not group results by country. Many seller profiles show a location, and every profile shows its pass and fail counts. We read all 466 profiles on 4 October 2026 and grouped them. In this guide, “listed in China” means the profile shows mainland China or Hong Kong.

Location on the seller’s Finnrick profile	Sellers	Published tests	Classifiable tests	Failed	Fail share
China, including Hong Kong	139	5,329	5,275	1,398	26.5%
United States	77	2,654	2,636	944	35.8%
Other countries	5	134	134	64	47.8%
No location shown	245	3,882	3,757	1,165	31.0%
All sellers	466	11,999	11,802	3,571	30.3%

Sellers listed in China supplied 44% of all published tests. The overall fail share fits Finnrick’s February 2026 statement that “roughly one in three fail basic expectations”.

Part of the gap comes from product mix. Retatrutide, tirzepatide and GHK-Cu are the three products that fail least often. They make up 72% of the classifiable tests from sellers listed in China, against 53% for sellers listed in the US. So the next step is to compare product by product.

High quality peptides from China: results product by product

The table covers every product with more than 100 published tests. Fail share is failed tests divided by passed plus failed. The figure in brackets is the number of classifiable tests behind it. The last column counts how many of the three public top places on that product’s Finnrick leaderboard, read on 4 October 2026, went to sellers listed in China.

Product	All sellers	Sellers listed in China	Sellers listed in the US	Top 3 places held by sellers listed in China
Retatrutide	22% (4,299)	20% (1,968)	26% (774)	1 of 3
Tirzepatide	22% (2,707)	20% (1,311)	23% (468)	1 of 3
GHK-Cu	13% (929)	13% (505)	9% (152)	2 of 3
BPC-157	51% (894)	45% (340)	57% (299)	2 of 3
Tesamorelin	53% (696)	49% (374)	55% (92)	2 of 3
Ipamorelin	32% (438)	25% (166)	35% (154)	1 of 3
Semaglutide	32% (369)	40% (67)	25% (207)	0 of 3
CJC-1295	71% (332)	51% (78)	80% (157)	2 of 3
TB-500	63% (278)	62% (143)	71% (56)	3 of 3
Cagrilintide	46% (193)	43% (96)	46% (13)	0 of 3
Melanotan II	38% (177)	22% (45)	49% (104)	1 of 3
PT-141	36% (148)	27% (48)	37% (68)	1 of 3

What the table shows:

• Lower fail shares in 10 of 12 products. Sellers listed in China failed less often than sellers listed in the US on every product except GHK-Cu and semaglutide. Cells with fewer than about 100 tests can move a lot on a few results.
• 16 of the 36 top-three places. Sellers listed in China held 16 places, and first place on 5 of the 12 boards. Seven of the 29 sellers in those places show no location, so the real count may be higher.
• A top place is relative. First place on the TB-500 board was rated 51%, and first place on the CJC-1295 board 49%. A top place means a seller did better than its peers on that product. Ratings are weighted towards recent tests, and thin histories are pulled towards the product’s average.

For the reasons some products fail more than others, see which peptides fail most often.

Research grade retatrutide from China: the most tested case

Retatrutide is the most tested product on Finnrick: 4,367 published tests from 378 sellers. Sellers listed in China supplied 1,968 of the 4,299 classifiable results, 46%, and failed 20% of them. One of them held a top-three place, rated 91% from 106 tests.

A 20% fail share still means one classifiable sample in five failed. For a GLP-1 sold under a research label, the first questions are whether the vial holds retatrutide and how much. Identity by LC-MS compares the measured mass with the expected mass. Content per vial gives the mg against a reference standard. When the label itself is in doubt, blind GLP-1 identification compares the masses in the vial with each candidate GLP-1 peptide, without being told the label. See the retatrutide page and the guide to retatrutide testing.

Why the seller’s location does not predict your vial

The figures above describe published samples. Five limits apply before you read them as a verdict on a country:

• The location field is incomplete. 245 of the 466 sellers, with 3,882 tests, show no location. The field records where the seller is listed, not where a batch was made. A seller listed in one country can sell peptide made in another.
• Buyers choose what to send. Samples come from anonymous public submissions and from Finnrick’s own purchases. They are not a random sample of the market.
• Some samples come from the seller. In Finnrick’s paid Launch programme, a seller sends four samples and two are tested. The results are marked “vendor-submitted”, and the seller can review and approve them before they are published.
• One lab ran most tests. Krause Analytical ran 7,328 of the 11,999 published tests, about 61%. Labs differ on mg; see why labs disagree.
• The default standard measures three things. Purity Focus checks identity, purity and fill. It does not grade endotoxin or sterility, so a pass says nothing about either.

Within each group, sellers differ widely. Among the 58 sellers listed in China with at least 20 classifiable tests, the share of tests that passed ranged from 42% to 94%, with a median of 73%. A seller’s record across many batches tells you more than its country. A result on your own batch tells you more again.

What the FDA found at GLP-1 ingredient makers

Regulators look at a different market: active ingredients sold to make medicines. The US FDA’s Import Alert 66-80, dated 21 September 2026 on the FDA site, covers GLP-1 bulk drug substances such as semaglutide and tirzepatide. It states that during the shortages, the ingredients used to compound GLP-1 medicines in the US came from makers around the world, including China, India and Europe.

FDA evaluated 48 GLP-1 ingredient sites, by inspection or by requests for records. It found 21% of them in breach of section 501 of the US drug law: their records showed failures of current good manufacturing practice, or they did not answer a records request in time. The alert does not break the 48 sites down by country.

A research peptide is not sold as a licensed medicine, and this alert does not test or grade research vials. It shows that a regulator found problems at about one registered site in five, and that it judges makers site by site.

Research grade peptides from China: what a grade label means

Listings use words such as “research grade” or “high purity”, or a grade with a number. These are labels the seller chooses. None of them names a method, a limit or a lab. Each claim on a label maps to a measurement you can ask for:

Claim on the listing	What a report should show
The peptide’s name	Identity by LC-MS: observed mass matched to the expected mass
“99% purity”	HPLC purity as a share of UV peak area, with the chromatogram and wavelength
“10 mg”	Content per vial in mg against a reference standard, with the basis stated (net peptide or total powder)
“Low endotoxin”	Endotoxin by LAL, as a number in EU per mg or per vial
“Sterile”	Sterility (USP <71>, 14 days) on unopened vials. A PCR screen or a microbial count is not a sterility test

Purity alone does not decide quality. Under Finnrick’s default standard, a sample at 99.4% purity fails even when it is the right peptide in the right amount, and a vial at 99.9% can still hold far less than its label. The guide to what peptide purity means explains the number.

Reputable Chinese peptide vendors and labs: how to judge one

A lab result describes vials, not a seller’s reputation, and this guide names no sellers. These checks work on any seller, wherever it is:

1. A batch report from a named outside lab. The lot on the report matches the lot on your vial.
2. The report opens at the lab’s own site. Type the lab’s address yourself and look the report up by its ID.
3. Raw data. A chromatogram and a mass spectrum, not only a table of numbers.
4. A long record. A rating built on 100 tests over two years says more than one built on 3. Of the 139 sellers listed in China on Finnrick, 62 had fewer than 10 classifiable tests.
5. Your own result. Random, sealed vials from your order, tested for identity, content and purity.

“Reputable Chinese peptide labs” usually means a manufacturer’s own quality control lab. It issues a COA for each batch, which is the factory testing its own product. In our September 2026 survey of labs that test peptides for this market, we found no independent lab in mainland China that markets testing to buyers abroad. The guides to checking a Chinese manufacturer’s COA and vetting a manufacturer go through each step, and peptide testing in China compares the options for an outside test.

What HK Analytical does

Our lab is in Hong Kong. Samples from mainland China reach it by cross-border courier: declare them accurately, follow the courier’s rules, and we send the sample form on request. One order can combine identity (LC-MS), purity (HPLC), content per vial, blind GLP-1 identification, endotoxin (LAL), Sterility (USP <71>, 14 days), residual solvents, counter-ion, water content and vial-to-vial consistency, in any mix per sample. 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.

Every number comes with its raw data. The report carries our photo of the vial as received, states how the client is related to HK Analytical, and is a signed record you can check on hkanalytical.com/verify. Like any lab result, it describes the vials tested and does not say whether a product is safe to use. To test a batch, request a quote or see the full test list.