Articles & News

FDA's Regulatory Procedures Manual calls a warning letter informal and advisory: the agency's position and a request to correct, not a court finding or a penalty. Read against two letters to peptide storefronts, with the reply clock, the close-out rules and the database's two dates.

Freeze-dried peptide takes up water from the air, and that water is weighed with the peptide. Karl Fischer titration is the test specific to water. Of 126 supplier records in this publication's ledger, none notes a certificate that reports a water content figure.

The small peaks beside a peptide's main peak are related substances: short sequences, oxidised and deamidated forms, and isomers. FDA draft guidance for generic peptides asks for each one above 0.5 percent to be identified. Of 126 supplier records in this publication's ledger, none notes a certificate that names one.

Synthetic peptides usually leave the lab as salts, and the counterion is weighed with the peptide: in a 2025 methods paper, TFA reached up to 35 percent of a test peptide salt's weight. Of 126 supplier records in this publication's ledger, none notes a certificate that reports a counterion amount.

HPLC purity is a share of the peaks a detector saw. Net peptide content is the share of the weighed powder that is peptide, after counterions and water. Published methods papers show the gap can be large, and a certificate that reports only purity has not measured it.

A US buyer working from a storefront they have never used can check four things before paying, and the order they check them in matters. This piece sets that order out and points to the specific findings, already published on this site, that support each step.

The FDA says Empower Pharmacy compounded essentially copies of approved semaglutide and tirzepatide products in volume, on prescriber notes that appear repeated word for word, and that its sterile-practice replies left gaps.

The only document a research peptide buyer usually holds is a third-party laboratory report, and the one line on it that would say who made the powder is filled in by whoever submitted the sample. Across this publication's ledger that line names the storefront itself, reads n/a or Unknown, or names an entirely different business.

This publication's own field guide to chromatograms explains why a peak's retention time is a fact about the method, not the compound. Three certificates in this ledger, all for the same 10mg retatrutide vial, show what that means in practice: the same compound, read by three different methods, produces three retention times that share nothing in common.

A certificate can be checked in about five minutes using three questions in a fixed order, none of which requires the lab's own portal: does the named laboratory exist independently, do the document's own dates make sense in sequence, and does the same figure repeat suspiciously across unrelated products. Two ledger cases show what each check catches.

A domain's registration record is a public document, and reading it takes about as long as checking a certificate. This piece shows one case where domain history cleared a suspected connection between two similarly named sites, and one where a young domain and a redacted registrant sat beside marketing built around a competitor's closure.

A supplier can sit in this registry for months without ever appearing in a rank, because observation and judgment are two different tiers here and the second is never built from the first alone. This piece reads the tier structure directly from its own documentation, and the rule that keeps a rank from outrunning its evidence.

A search engine ranks storefronts by marketing spend. This registry is built from a domain list instead, screened by a fixed, repeatable check anyone could run by hand. This piece reads that check directly from its own source, including a pilot run that found the wrong reading order, not the list itself, was producing almost nothing useful.

This publication's own review rubric names three forms a sentence about a supplier can take, and states plainly that only one of them is checkable by a reader. This piece reads that rule directly from the source, and shows the test any reader can run on any directory that writes about named businesses.

A certificate is only as strong as the laboratory behind it, and this ledger names the same handful of laboratories over and over. This piece reads four of them closely: the two most frequently named, one holding a real accreditation for the wrong kind of testing, and one carrying an accreditation a vendor's own page attributes to it that its own site never claims.

Third-party tested is not a standard, a certification, or a claim with one fixed meaning. This piece reads three ledger cases spanning the range: a vendor whose central quality claim names no laboratory at all, one whose named testing partner matches the documents on some product pages and not others, and one running a browsable, batch-specific record a buyer can check without asking.

Fifty of the 126 supplier records in this publication's ledger mention a certificate library or index by name, a structural claim distinct from posting one document per product or none at all. This piece reads two of those libraries closely: one that states its own size, and one that is missing documents the issuing laboratory actually holds.

A handful of pages on this site carry a disclosure paragraph above the first line, and no article's own front matter turns it on or off. This piece reads the mechanism that decides when the disclosure appears, the FTC standard behind it, and what a reader should do with a disclosure like this on any site.

Ninety-two of the 126 supplier records in this publication's ledger name BPC-157, and a same-day query of ClinicalTrials.gov's own registry returns exactly four studies of it, one of them a gummy supplement product rather than a trial of the peptide itself. This piece reads what each of the four actually says and what the registry does not cover.

A certificate can be genuine, current, and still describe nothing in particular if the lot number that ties it to one production run is missing, unmatched, or simply out of date with what a storefront's own index says is for sale. This piece reads the federal definition behind the field and four ledger cases where it worked, half-worked, or was left blank.

A certificate can pass every check for a real lab and a real batch and still be describing something that has nothing to do with the vial a buyer is about to receive. This piece explains the three dates on a certificate, what order they should fall in, and shows dated ledger cases where a genuine document quietly outlived the product it once described.

21 CFR 809.10(c), the regulation two peptide sellers in this publication's ledger cite for their research-use-only label, governs in vitro diagnostic products: reagents and instruments used to diagnose disease from a human specimen. The FDA guidance that explains it never mentions a research chemical sold in a vial, and FDA reads what a seller actually does, not what the label says.

A certificate that reports purity has not reported potency, and a certificate that reports either has not reported identity. Four of 126 supplier records in this ledger carry a distinct potency assay at all, and reading the difference is what tells a reader how much of an over-label vial is peptide.

Certificate access follows three patterns in this ledger: published openly, released only after registration, and issued on request. One case shows the comparison most sharply, a gated storefront and a second domain carrying the same brand name, whose relationship to it is unconfirmed, publishing the same kind of document without a login.

Peptide Sciences' own domain returned nothing but a voluntary closure notice on every path checked on 2026-08-20, and its Trustpilot profile is among those whose address returned the platform's removal notice rather than a profile on 2026-08-25. Five other domains carry the name, and this piece states plainly what has and has not been checked about each.

Eight terms that appear on nearly every supplier page in this ledger, defined plainly, each ending with the one question the term actually answers and the ones it does not.

Three short emails a researcher can send before paying: one asking for the batch-specific certificate rather than a catalog figure, one asking for the report number a laboratory portal can verify, and one asking plainly what the certificate does not test. Each is grounded in a documented case where asking, or not asking, changed what a buyer would have known.

Across 53 price records from 50 suppliers pricing the same benchmark, a single 10mg retatrutide vial ranges from under $9 to $200, a spread of more than twenty times. The documentation behind the price, not the price alone, is what tells a buyer which end they are looking at.

A supplier badge naming ISO 17025, cGMP or USP grade is a claim about a specific, narrow thing, and most of them are checkable in a public registry in a few minutes. This explains what each claim covers, what it does not, and how the ledger's own supplier records show the difference between a checked claim and a bare one.

An intervention search of clinicaltrials.gov on 2026-09-16 returned 44 registered studies for cagrilintide and 18 for eloralintide, and PubMed held 99 records against 9. The pharmacology papers put the two molecules in different places on the same receptor family.

Payment method says nothing about what is in a vial and a great deal about what a buyer can do afterward. This explains which federal rule attaches to which rail, who screened the seller before the card field appeared, and what the storefronts in this publication's ledger charge for the difference.

Each of the six flags on this index is something a storefront publishes about itself, and each one takes away a check a buyer could otherwise run. The index explains what every flag removes, with the regulator or standards body that defines the missing check.

Forged certificates of analysis are no longer a rumor in this market. A federal sentencing in July 2026 drew a news account reporting that the seller admitted his business forged them, and the checks that catch one are the laboratory, the report number, the internal dates, and whether the figure describes a lot or a whole catalog.

A chromatogram is a picture of one method's field of view. Peaks are things that left the column and absorbed at one wavelength, area percent is arithmetic over those peaks, and a related-substance peak stays unnamed until a mass spectrometer names it.

A research peptide is a short chain of amino acids, short enough to be built by chemistry on a resin rather than grown in cells, and sold under a research-use-only label. That label is set by a labeling regulation for diagnostic products and explained in a 2013 FDA guidance, and neither defines anything for a vial of powder. What separates suppliers is the paperwork around a lot: a certificate, a named laboratory, a method, and a lot number that ties the vial to the document.

A stability-indicating HPLC method is one validated by deliberately damaging the peptide and showing the method still separates the intact molecule from the damage. A 2026 paper in BMC Chemistry did that for semaglutide and tirzepatide under five stress conditions, with 11.84 to 18.97 percent degradation and a resolution of at least 2 wherever a degradant peak appeared. No record in this publication's ledger says whether a certificate's purity method was ever tested that way.

Five storefronts wrote the research-use disclaimer five different ways. The sentence the agency wrote next is identical in all five. Four of the letters make the diluent a cited drug of its own; the fifth points instead at a calculator and a guide.

Four newer suppliers ranked the same day on the same checks, from certificate practice to refund clarity. A few match or beat the older houses, and each entry says what could be verified and what could not.

Of 125 suppliers with a shipping record, 14 publish nothing a visitor can read and 24 state their own terms two different ways. What each kind of promise commits to, what a tracking number does not, and the federal floor under all of it.

Six suppliers ranked only on the paperwork a stranger can check: batch-specific certificates, named labs, and purity given with its method. Price and catalog did not move a single rank.

Six suppliers that stock genuinely rare research compounds, ranked on how far the catalog reaches and whether a batch certificate keeps up with it. The biggest catalog does not win.

Certificate practice weighted first, then pricing honesty, catalog and shipping terms. Six suppliers, every claim traced to a dated record or a named third-party source.

Six suppliers ranked on the prices actually recorded on their pages, weighed against the certificates behind them. A low sticker with no checkable paperwork loses to a fair one that has it.

Forty-four of the 126 suppliers in this ledger have endotoxin somewhere in their record. Eighteen ever printed a figure, in four units that do not convert into each other.

A laboratory is never accredited in general. It is accredited for specific methods written on a scope document anyone can pull. Checking that document is the whole of the check.

On 2026-08-25 we opened a Trustpilot page for all 59 suppliers in our ledger: 21 removal notices, 11 that never carried a profile, 27 live. What that leaves.

Seven FDA warning letters to online peptide sellers, all dated March 31. The agency read each seller's own product pages as the evidence, and the bacteriostatic water sold beside the peptides became a drug on its own.

Why research peptides ship as a dry solid, what removing the water actually buys, and why a popular storage rule of thumb holds up worse than it sounds.

A plain walkthrough of a Certificate of Analysis. What each number measures, what it can back up, and how to check the lab that issued it.

What a purity number from an HPLC run actually measures. One chromatogram, one method, one day, and identity is a separate question.

We walked 21 supplier checkouts and stopped at the payment screen. Many take no card at all, and the methods that replace cards cannot be charged back.

Supplier sites rarely vanish overnight. They signal it for weeks first, usually starting when card payments quietly disappear. Here is what to watch before you pay.

A certificate is only as strong as the lab behind it. How to find out whether that lab is real, accredited, and willing to answer.

The small amount of math behind a research solution. One equation, a few unit conversions, and the milligram-and-microgram slip that can throw a concentration off by a thousandfold.

Reading a certificate is not the same as checking it. The big testing labs run public portals, and matching a report number takes under two minutes.

A purity percentage measures one thing by one method. What the number covers, what it leaves out, and why two honest vials can still differ.

A checklist for sizing up a supplier in about twenty minutes. Start with the certificate, because it fails fastest and tells you the most.

Purity and identity are different questions. HPLC answers the first, mass spectrometry answers the second, and most certificates stop at one.

A standard certificate covers purity and sometimes identity. Sterility, endotoxin and water content are separate tests, and a certificate that skips them is not a clean sample, only a question no one asked.