Informational only. Not medical advice.INFORMATIONAL PLATFORM ONLY — NOT MEDICAL ADVICE, DIAGNOSIS, OR TREATMENT
Informational only. Not medical advice.INFORMATIONAL PLATFORM ONLY — NOT MEDICAL ADVICE, DIAGNOSIS, OR TREATMENT
We analyzed 5,232 Certificates of Analysis published by 130 research-peptide vendors and tested by 39 labs, all sold Research Use Only. The purity is real. The harder contamination tests — sterility, endotoxin, heavy metals — mostly aren’t published.
21.4%
of lyophilized COAs publish a sterility result
56.5%
of vials land within ±10% of their label
98.1%
clear 98% HPLC purity
8
clusters of storefronts sharing a physical batch
Of the 4,808 lyophilized-powder peptide COAs in the corpus, only about one in five publishes a sterility result and one in five publishes both sterility and bacterial endotoxin. These are the harder, costlier contamination tests: HPLC purity confirms the molecule matches the label, but says nothing about microbial or endotoxin contamination — and endotoxin contamination alone can invalidate the cell-culture and in-vivo research these materials are sold for.
Across 3,560 COAs that report both a measured net-peptide content and a labeled amount, only 4.4% come in more than 10% under label. The far more common miss is the other direction: 38.9% run more than 10% over. Only 56.5% land inside a tight ±10% of what the label claims.
Overfilling isn’t generosity — it points at fill processes tuned to clear a minimum-content check rather than hit a spec. For the buyer it means the milligrams in the vial are a moving target from batch to batch.
This isn’t a fear campaign. Of 4,506 COAs carrying an HPLC purity result, 98.1% clear 98% and only 0.5% fall below 95%. The chemistry is largely solid; the gap is in the tests vendors don’trun and the labels they don’t hit.
At least 8 clusters of separately-branded vendors share a physical batch— the identical lot number printed on COAs sold under different names — or a single street address. Shared labs or platforms don’t count here; only signals that are hard to fake. A sample:
Arcane Peptides ↔ Moglabs
19 shared lot numbers across NAD+, Semax, Retatrutide, GHK-Cu, PT-141, and more
Orbitrex Peptides ↔ Synthesis Peptides
11 shared lot numbers across NAD+, BPC-157, Retatrutide, SS-31
Hydro Research Peptides ↔ Optima Supply
shared lots across Survodutide, Mazdutide, Cagrilintide/Semaglutide, BPC-157
Simple Peptide ↔ Alpha Omega ↔ EZ Peptides
shared lots + one Delray Beach street address
The live version of this analysis runs on the Forgery Map.
Snapshot taken July 16, 2026 from the Disclosed Labs corpus: 5,232 Certificates of Analysis that were successfully machine-extracted, in-scope (peptides, not steroids/SARMs), currently published, and not withdrawn. Testing-coverage figures count whether a COA publishes a given result — the observable fact — not whether a vendor privately ran the test. Label-accuracy figures use only COAs reporting both a measured net-peptide content and a labeled milligram amount; multi-peptide blends, whose per-component labels understate summed content, were reviewed and excluded from the extreme tail. Operator clusters are built only from hard-to-fake shared signals (identical COA file, identical lot number, phone, or street address — excluding shared labs and registered-agent mail drops).
The products analyzed are sold Research Use Only. This study characterizes the testing published on COAs — sterility and endotoxin are contamination-characterization tests relevant to research validity. Nothing here describes, recommends, or implies human use, dosing, efficacy, or route of administration. Every figure is reproducible against the public corpus.
Cite this study
Disclosed Labs. “The State of Peptide Testing 2026.” July 16, 2026. https://www.disclosedlabs.com/research/peptide-testing-2026
Disclosed Labs