The Peptide Quality Crisis: What Investigations Found and How to Protect Yourself

In the first two weeks of April 2026, five major outlets published peptide investigations: ProPublica, The New Yorker, The Guardian, NBC News, and CBC Radio. None of them coordinated. The market had outrun its oversight, and the reporting all caught up at once.

ProPublica looked at the FDA's plan to bring 14 restricted peptides back into legal compounding, and what that would mean for products that haven't been tested seriously. Dhruv Khullar's New Yorker piece followed peptides out of CrossFit gyms and into mainstream wellness clinics, with lab tests showing contamination along the way. The Guardian put a number on it — about a third of peptide products fail basic quality checks. NBC News asked whether the science is keeping up with demand (it isn't). CBC Radio was blunter and called injectable peptides "a giant scam."

Three countries' regulators moved in the same window. The UK's MHRA opened investigations into peptide clinics making medicinal claims for unregulated products. Health Canada published an advisory naming BPC-157, CJC-1295, and retatrutide among seized injectables. A federal grand jury in Utah indicted an osteopathic physician for importing misbranded Chinese peptides and selling them to more than 200 patients.

The number everyone is quoting comes from Finnrick Analytics, an independent peptide testing lab in Texas. Finnrick has tested more than 6,600 samples from over 200 vendors across 15 popular peptides. Roughly 30% are mislabeled, under- or overdosed, or contaminated with toxins or bacteria. The New Yorker and The Guardian both ran with the figure.

Finnrick now processes around 60,000 samples a year. A few years ago they were processing a handful a month. Their public vendor ratings show enormous spread: some come back near-perfect, others fail consistently on purity, potency, or sterility.

Worth being clear about what the 30% covers and what it doesn't. Finnrick's pool includes everything from FDA-registered 503B outsourcing facilities down to gray-market "research chemical" sites selling vials labeled "not for human use." Failure rates at the gray-market end are almost certainly worse than 30%. At licensed compounders they are better — though not by as much as you might hope. The FDA's own historical testing of compounded products has found potency failures in 25-33% of samples.

The contamination is concrete. Five different things go wrong in the testing data, and they often go wrong together.

Plain potency failure is the most common. A vial labeled "5mg BPC-157" might contain 3mg, 7mg, or nothing at all. Underdose and the peptide doesn't do what it's supposed to. Overdose and the side effects get unpredictable. Zero active ingredient means whatever else is in the vial — solvent, filler, contamination — is what's going into your tissue.

Endotoxins are the dangerous one in the short term. They are bacterial cell-wall fragments, and they trigger a strong immune response. Injectable products are supposed to stay below the limits in USP <85>; anything compounded without real sterile technique can blow past them. The clinical picture is fever, chills, falling blood pressure. In bad cases, septic shock.

Heavy metals show up because of impure starting material and dirty manufacturing equipment. The New Yorker found lead specifically in BPC-157 samples. ICP-MS testing catches metals down to parts-per-billion, and most gray-market sellers don't bother running it.

Microbial contamination is the constant risk for any injectable made outside a sterile environment. USP <61> and <71> are standard pharmaceutical tests, and rarely done on research-grade peptides.

Mislabeling is the broad category that includes everything else: wrong peptide entirely, wrong concentration, missing excipient information, fabricated lot numbers that don't correspond to any real manufacturing batch.

Most peptide active ingredients — APIs in pharmacy terms — come out of a small group of Chinese chemical manufacturers. That's true at both ends of the market. The difference is which Chinese manufacturers, and what documentation comes with the product.

Licensed compounding pharmacies buy from FDA-registered API makers and get full paperwork along with the material: certificates of analysis, drug master files, cGMP records. The FDA publishes the registry of foreign API manufacturers it inspects.

Research chemical vendors source from Chinese manufacturers that aren't on that registry, working to whatever quality standard they choose. The Utah physician indicted in April 2026 was charged with buying tirzepatide, semaglutide, BPC-157, and retatrutide from a Chinese supplier, putting fake labels on them, and selling them to patients. Customs seized 5,000 units at the border.

Pricing reflects all of this. A 5mg vial of BPC-157 from a licensed compounder with full testing usually runs $50-100. The same vial from a research chemical site runs $15-30. The gap isn't manufacturing efficiency. The gap is that testing, documentation, and sterile fill weren't done.

All three regulatory responses landed in the same window in April 2026, even if no one was coordinating them.

In the US, the FDA sent warning letters to more than 30 telehealth companies for illegally marketing compounded GLP-1 products. The Department of Justice indicted the Utah osteopath. At the same time, the HHS Secretary's push to put 14 restricted peptides back into legal compounding created an awkward parallel: the same federal apparatus expanding legal access while prosecuting illegal distribution.

In the UK, the MHRA opened an investigation into peptide clinics charging patients hundreds of pounds a month for unregulated products and making medicinal claims about them. Under UK law, a therapeutic claim turns a product into a medicine, which then requires a marketing authorization. None of the clinics had one.

In Canada, Health Canada issued a public advisory against unauthorized injectable peptides and named BPC-157, CJC-1295, and retatrutide among seized products. The advisory cited liver and kidney damage, blood clots, and tumors. Some of those harms come from theoretical mechanisms more than documented cases in actual peptide users — worth noting before quoting Health Canada as if every risk on the list has been observed in patients.

What all three agencies share is a market that has outrun their enforcement capacity. NBC News pegged US peptide-related searches at 10.1 million in January 2026. Frameworks built for traditional pharmaceuticals don't fit a category whose distribution channel is research-chemical websites.

The clearest defense is knowing what a clean product actually comes with.

A real product has a Certificate of Analysis tied to a specific lot number, run by an accredited lab — ISO/IEC 17025:2017 is the accreditation marker to look for. A clean COA covers five tests:

For compounded peptides specifically, the pharmacy itself matters as much as the COA. It should be a state-licensed 503A or FDA-registered 503B. The API should come from an FDA-registered manufacturer. The product should be prepared sterile with a beyond-use date that makes sense. A pharmacy that won't send you the documentation when you ask is telling you something.

The red flags are mostly the inverse of all that. No COA. A COA from the manufacturer instead of an independent lab. A COA with no lot number, or a lot number that looks like a template. "For research use only" labels on something being sold for human use. A price far below the rest of the market. A vendor that won't say where its API came from.

One distinction most coverage skips: not all "compounded" peptides are the same thing. There are roughly three tiers, and the risk profile differs more between tiers than between peptides.

503B outsourcing facilities are the top of the legal pile, about 70 of them in the US. They are FDA-registered, have to follow cGMP, get inspected, must report adverse events, and can compound in bulk. Their products go through standardized quality testing.

503A pharmacies are a much bigger group, more than 50,000 of them, regulated by states rather than the FDA. They compound to individual prescriptions and aren't required to follow FDA cGMP. Quality varies a lot between them. If a 503A won't tell you about its testing protocols and where it sources API from, that's a sign.

Then there's the research chemical layer, which is less a tier than a separate market. Unregistered, no required testing, no adverse-event reporting. The "not for human consumption" label is doing legal cover work, not warning anyone. This is where the 30% number probably understates the problem.

The point isn't that peptides as a class are dangerous. It's that the source of any given vial tells you more about your risk than the molecule inside does.

April 2026 is probably the inflection point for how this market gets regulated. It can't keep growing in a vacuum, and the open question is what it grows into.

If reclassification works as intended, the 14 peptides moving back into legal compounding pull a chunk of demand off research-chemical sites and into 503A and 503B pharmacies. Independent testing — Finnrick and the labs coming up behind it — keeps scaling and creates the kind of transparency that pushes bad sellers out. If reclassification doesn't work, it just adds demand without adding supply quality: 503A pharmacies cut corners to keep up, the gray market holds its price advantage, and patients read "legal" as "tested" when those are different things. The second outcome currently has more momentum behind it.

If you're using peptides or thinking about starting, the actual advice hasn't changed much. It just carries more weight than it used to. Work with a clinician who actually knows what they're prescribing. Use a licensed compounder. Ask for a certificate of analysis from an independent accredited lab. A price that looks too good almost always is.

Peptides themselves aren't the problem. Several are FDA-approved with strong evidence behind them. The problem is the gap between what's printed on the vial and what's actually in it. That gap is the part worth being careful about.