Methodology

We read each vendor's own pages: home, product, COA or lab-results page, shipping, payment and policy pages. For each fact we store the exact page and a short quote. When we cannot find something — for example a public COA page — we say "not found" rather than guess.

Regulatory status comes from FDA's warning-letter database. A letter is linked to a vendor only when the letter names the vendor's own domain or company. See all peptide warning letters.

No vendor is starred or scored, and no seller can pay to appear, move up or be removed. The Peptide Directory's owners also operate businesses in the peptide market; see ownership.

The vendor directory has a default order that starts from published criteria that use only facts shown on the page; editors also adjust placement to spread out vendors with related ownership or shared suppliers, including businesses connected to our owners, so no group clusters together. The adjustment is a fixed spacing step applied by code after the criteria sort: one vendor from such a group is placed near the top (never first) and the others are spaced through the rest of the list, with other vendors between them; which one is placed near the top varies by list. Editors can also pin a vendor to a set position in the default order, which may be first. It can therefore move a vendor connected to our owners higher than the criteria alone would. Use the sort controls to see a strict ranking: the strict criteria ranking applies only the criteria below, with no adjustment. Each step of the criteria only breaks ties left by the step before it:

1. Transparency tier (2, 1 or 0). One point for a public certificate-of-analysis or test-results page we could open before buying, one point for naming an outside testing laboratory from our list of established peptide-testing labs (Janoshik, Kovera, Vanguard Laboratory, Chromate, Freedom Diagnostics, Mistral Diagnostics, BioRegen, TrustPointe, MZ Biolabs, Eurofins, Colmaric, Analytical Resource Laboratories) on the vendor's own pages. A lab we cannot independently verify (no checkable legal entity, address or accreditation) is shown on the vendor's page but does not earn the point; the same rule applies to every vendor.
2. Core compounds in stock (0–5). How many of the five compounds in our price dataset (BPC-157, TB-500, semaglutide, tirzepatide, retatrutide) the vendor had at least one in-stock listing for when we last checked. More first.
3. Typical price per mg. For each of those compounds we take the vendor's lowest in-stock price per mg (including 5- or 10-vial packs where the store sells them, because that is a price any buyer can pay) and divide it by the median for that compound across all vendors; the vendor's index is the median of those ratios (1.00 = typical, 0.50 = half the typical price). Lower first. Vendors with no in-stock tracked compound come after those with one.
4. Name, A–Z, as the last tie-break.

The order is not a quality ranking: a public COA page, a named lab and a low price do not show that any product is pure, sterile or correctly dosed, and we have not tested products. Each vendor's tier, stock count and price index are printed in the table under the list so the strict ranking can be checked. FDA warning letters are shown next to every vendor but are not part of the sort. Prefer plain alphabetical order? Use the A–Z view.

Prices are read from each vendor's public store data (Shopify product JSON, the WooCommerce Store API or the product page's structured data) and linked to the product page. Multi-vial packs are converted to total milligrams; blends, capsules and kits are excluded from per-mg comparisons. Last checked October 19, 2026. Prices change; the vendor's page is authoritative. Price tracker · CSV.

Identity data comes from PubChem (only when a PubChem synonym matches the compound name), research counts and citations from PubMed, trials from ClinicalTrials.gov and approvals from Drugs@FDA. Each citation's PMID is resolved against PubMed before it is published. Compounds with too little public data are marked as limited and kept out of search indexes.

Clinic listings start from public business listings. We then read each clinic's own website and list a service only when the site mentions it, keeping the page and quote as evidence. Prescriber names are cross-checked against the NPPES NPI registry where possible. Sites using "research use only" language are excluded.

We publish doses only as they appear in FDA labels or published trial protocols, with the source. Reconstitution pages are arithmetic, not advice. Nothing here is medical advice.

Found an error? Tell us through the contact page; fixes are logged on the corrections page.