The Beginning of the End for RUO

In the last three weeks, the research-use-only peptide market lost the argument it was built on.

On August 12 Eli Lilly sued four RUO peptide vendors by name. On August 21 FDA revised the import alert that holds GLP-1 API at the border. On August 24 FDA signed five warning letters to research-peptide sellers, and posted them to the public database on September 1. On August 27 the Fifth Circuit closed the last appeal keeping the compounding fight alive.

None of those is a ban. Together they are something more specific and more dangerous to a seller: a coordinated attack on the label itself. FDA now says on the record that "research use only" does not save a product from being an unapproved new drug. Lilly's lawsuits go further and treat the phrase as the pretext.

So here is the reptides position, dated September 1, 2026, written down so it can be checked later.

One honest caveat before the evidence, because it is the strongest argument against everything below. Moving to a novel sequence gets a vendor out from under Lilly. It does not get them out from under FDA. That distinction runs through this whole piece.

Four of those six are research-use-only peptide vendors. That is the part to sit with. Lilly has sued compounders and telehealth firms before. Naming RUO sellers as defendants is new.

How the suits are pleaded is the strategy. Five of the six run on state unfair-competition and consumer-protection statutes under diversity jurisdiction. Only the med spa case carries a federal Lanham Act claim. Lilly is not suing the vendors on trademark, because there is no retatrutide trademark to assert. There is no approved retatrutide product at all.

What Lilly argues instead is that "research use only" is a pretext. The complaints allege the product is "not a chemical intended for research use but instead a drug it intends and sells for human use," and plead that as an unfair or deceptive trade practice premised on selling an unapproved new drug. The Texas Peptides complaint alone runs eight counts, under the consumer-protection statutes of Alaska, Colorado, Connecticut, North Carolina, South Carolina, Tennessee and Washington plus Texas common-law unfair competition, with no federal claim at all.

Read the structure rather than the headline. The label is not being attacked as a standalone false statement. It is being treated as the cover story on top of the real allegation, which is that these are unapproved drugs sold to people.

The same day, Lilly said it had referred more than 200 individuals and entities to FDA, the Department of Justice, state attorneys general and law enforcement, and reported more than 14,000 websites, advertisements and listings across more than 100 countries. It publicly called on online platforms, credit card companies, payment processors and shipping carriers to cut the sellers off.

Lilly associate vice president Max Denning called the problem global and enormous and said plainly that six lawsuits will not solve it. Read that as wave one. It works on contact: three of the sued vendors pulled their retatrutide listings around the day the complaints were filed, before answering anything.

Be accurate about where the cases stand, though. As of September 1 there is no injunction, no default, no settlement and no damages award in any of the six. The Striker docket shows a complaint, a summons issued August 17, a waiver of service executed August 25 and a certificate of interested parties on August 28. Anyone telling you a court has ruled on the RUO label is ahead of the record.

The revision added orforglipron API to the Green List and attached new product codes covering "Hormone N.E.C." categories alongside the existing "Peptide N.E.C." codes. Widening the codes is the quiet part. An alert written around GLP-1 API catches GLP-1 API. An alert with hormone and peptide catch-all codes attached catches a great deal more of a parcel manifest.

The products named across those letters are the whole shelf: tirzepatide, semaglutide, retatrutide, survodutide, mazdutide, SS-31, PT-141, tesamorelin and bacteriostatic water. Two things in the letters matter more than the product names.

First, FDA rejected the disclaimer in writing. The NuScience letter notes that the firm labels products "for laboratory, research, and analytical use" and "not intended for human or veterinary use," then holds that the website evidence "establishes that your products are intended to be drugs for human use." The label does not decide the question. The marketing does.

Second, FDA is treating the accessories as evidence. The NuScience letter cites the firm's peptide calculator. The Peak Performance, Peptide Partners and Tex Peptides letters cite bacteriostatic water sold as a solvent for reconstituting the products, which FDA says provides the means to prepare an injectable drug for human administration.

It is not the first time either. FDA issued seven letters dated March 31, 2026 under the same heading, including to Gram Peptides, Prime Sciences and Mile High Compounds, and reached the same conclusion about the label. The federal position was five months old before Lilly filed anything.

August did not come out of nowhere. The two years before it read as a ratchet, and each click removed an excuse.

The shortage excuse went first. For tirzepatide, the 503A enforcement-discretion period ended after the district court denied a preliminary injunction on March 5, 2025, with 503B running to March 19. For semaglutide, 503A ran to April 22, 2025 and 503B to May 22. Neither molecule now appears on the 503B bulks list or on FDA's shortage list.

The border excuse went next. FDA established a GLP-1 API green list on September 5, 2025, then issued Import Alert 66-80 on September 19. Then the volume arrived. Customs officers at the Port of Cincinnati intercepted roughly 5,000 individual peptide shipments smuggled from China inside more than 300 mis-manifested master cartons. By Lilly's own count, CBP made over 1,400 seizures of illicit GLP-1 drugs totalling nearly 90,000 vials in July 2026 alone, against more than 690 shipments and 31,000 units in all of fiscal 2025. One month of 2026 is roughly three times the annual seizure volume of 2025.

The "nobody actually gets prosecuted" excuse went in July. On July 30, 2026 Matthew Kawa, owner of Paradigm Peptides, was sentenced to 70 months in federal prison and agreed to forfeit $5 million. Prosecutors estimated Paradigm had about 54,000 unique customers across all 50 states and 80 countries. The company forged laboratory certificates of analysis, imported from China and India while marketing the product as American-made, and sold six compounds marketed as testosterone mimics that government testing found actually contained testosterone.

The forged certificates are the detail worth carrying out of that case. The COA is the entire trust object in this market, and 70 months is the first real price anyone has paid for faking one.

The counterweight is real. Ohio's boards have summarily suspended more than thirty clinics and med spas since the start of 2025, but a Washington Post investigation published July 24, 2026 surveyed state pharmacy, medical and nursing boards and found most states lack the authority or the resources to oversee peptide sales through med spas at all, concluding those businesses face less inspection than restaurants. Federal pressure is rising against a state-level floor that is close to absent.

The market all of this is aimed at got very large, very fast, and the growth is measurable from outside.

Chainalysis put crypto inflows to gray-market peptide vendors at $32 million in Q1 2026, up 159% quarter over quarter from $12 million, with Q2 pacing at $39 million and the ecosystem on track to cross a $100 million annual run rate. That is against roughly $1 million per quarter through 2024, six consecutive quarters of growth, with a step change from about $200,000 a month before 2025 to nearly $10 million a month once looksmaxxing content went viral on TikTok.

That is only the crypto slice. Bloomberg Businessweek's July 2026 feature put the whole black market at $1 billion to $3 billion, with the Czech testing lab Janoshik Analytical at roughly $100 million in annual revenue and test volume up more than 1,200% from 2023 to 2025. Upstream matches: US imports of hormone and peptide compounds from China roughly doubled to $328 million in the first three quarters of 2025.

Now the price, because a common claim in this market is that gray peptide prices have collapsed. We went looking for a defensible time series and could not find one, so we are not asserting it. What is documented is the gap. In February 2026 a Chinese seller offered tirzepatide at $48 for ten 5 mg vials, about $0.96 per mg, plus $60 shipping and $20 customs, near $50 a month against Zepbound at $399. Yale modeled injectable semaglutide manufacture at $0.89 to $4.73 a month against a Wegovy list around $1,349. Legendary Peptides, one of the sued vendors, listed retatrutide from $75 to $210.

So the froth is not cheapness. It is volume, velocity and marketing running ahead of verification. The most damning number in the Chainalysis work is not the revenue. Per-buyer spending on third-party purity testing fell roughly 88%, to about $8 per buyer, while total testing volume rose. More people are buying, and each of them is checking less. What that does to quality is not a matter of opinion.

• Finnrick, a Texas peptide testing lab, told the Guardian in April 2026 that about one third of the thousands of gray-market products it analysed failed basic quality checks, across identity, purity and quantity.
• FDA's January 20, 2026 warning letter to the outsourcing facility GenoGenix documents an unopened NAD+ vial carrying 3,360 EU/mL of bacterial endotoxin, after three patients dosed from that lot were sent to the emergency room.
• A 2024 analysis of no-prescription semaglutide bought online measured purity of 7.7% to 14.37% against a claimed 99%, with endotoxin in every sample (Ashraf, J Med Internet Res 2024, PMID 39509151). A systematic impurity analysis of falsified polypeptide drugs found purity from about 5% to 75%, plus arsenic and lead above the parenteral limit in some samples (Janvier, Talanta 2018, PMID 30029448).
• A FAERS study of 81,078 GLP-1 reports found 707 involving compounded product, with sharply elevated reporting odds for contamination (ROR 19.00) and preparation errors (ROR 48.92).
• A Public Health Reports study published August 2026 asked 20 adults to self-administer gray-market semaglutide using only information they could find online. They completed a mean of 15 of 34 required steps correctly.

That last study is the froth in one sentence. The market solved distribution and it never solved competence.

The reptides market map has carried this call for a while, and it predates the August filings: sustained enforcement makes RUO retail more evasive rather than smaller. Vaguer copy, faster domain churn, overseas fulfillment, more reliance on private communities.

The documented pattern here is adaptation. Import alerts, closed compounding windows, 5,000 seized shipments and the March 6, 2026 voluntary closure of Peptide Sciences, the largest US gray-market vendor by traffic, did not end the channel. Its catalog was absorbed by competitors through mid-2026.

We do not have to guess at the next step, because FDA already documented it. In the March 31, 2026 letters, Gram Peptides was listing retatrutide as "GLP-1-R peptide" and tirzepatide as "GLP-2 peptide," Prime Sciences was selling "GLP1-R," "GLP1-S" and "GLP1-T," and Mile High Compounds was marketing "GLP-3 RT (Retatrutide)." Five months later the August 24 letters name the same trick: NuScience selling "GLP-2 Tirz Peptide" and "GLP-3 RT (Retatrutide) Research Peptide," Peak Performance selling "GLP-3R."

That is the correction we owe our own thesis. The 2026 evasion playbook so far is renaming, not re-sequencing. Vendors obfuscated the label and shipped the same patented molecule. It did not work, because FDA reads the whole website, and a code name that resolves to retatrutide in the product description resolves to retatrutide in the letter too.

Renaming is phase one because it is free. It buys a quarter of search-engine ambiguity and nothing else. What comes after it is the expensive move, and it is the one that actually changes the legal picture.

Sort this market by who can sue you and the future gets legible.

The third row is already underway and it is specific enough to name.

FLGR242 is the clearest example, because it shows what this next phase actually breaks. It is a follistatin-344 derivative sold under that same vendor code by at least six independent RUO sellers. We confirmed live listings at Peptagon (10 mg, two vials, $631, sold out), Molecular Edge Peptides, Paramount Peptides ($499 for two 5 mg vials), Sanavital ($1,500), Synergistic Nutrition and BioLongevity Labs, whose 10 mg pair lists at $499.00 and up, discounted to $249.50 at the time of writing. There is no pharma originator. No INN. No approved comparator. No patent holder with standing to enjoin anyone.

There is also no agreement on what it does. BioLongevity's copy says FLGR242 is "a novel fragmented, modified version of Follistatin-344 (FST-344) that does not bind to the protein activin." Sanavital says it "potently inhibits myostatin and activin." A reference page describing the same compound says it "retains full binding activity against its primary research targets: myostatin and activin."

That is not a quality-control problem. It is the absence of the thing quality control checks against.

It is not alone. MID-35, a myostatin inhibitory peptide first reported out of Kyoto Pharmaceutical University in 2022, is supplied by RUO vendors for laboratory research. SLU-PP-332 and SLU-PP-915, estrogen-related-receptor agonists marketed as exercise mimetics, are listed by peptide vendors despite not being peptides at all. The channel is broadening past peptides into whatever the preclinical literature offers, which is exactly how research-chemical vendors diversified.

And the analogy is not a metaphor. Chainalysis traced two suppliers now serving the retail gray-market peptide trade, Shanghai Sigma Audley and Bigreat Technology, back to selling fentanyl and amphetamine precursors to transnational drug networks. Same suppliers. Same crypto rails. Same "not for human consumption" framing. A former quant trader told Fortune the market was "Silk Road all over again," after buying 5-amino-1MQ for $49, paid in USDC, from a vendor traced to a Shanghai firm that took in $3.6 million between late January and early June 2026.

Here is what that costs a buyer. A compound like FLGR242 has no published human data, no toxicology, no pharmacokinetics, no immunogenicity work and no approved reference standard. And when a vendor describes an analogue using research performed on the parent protein, the citation is doing work the molecule has not earned. Follistatin-344 research is not FLGR242 research.

The evidence layer reptides grades on, registered trials, PMIDs, approved labels, registry entries, does not exist for a molecule with no history. We grade evidence. A compound with no evidence does not get a low grade. It gets no grade, and no grade is the finding.

The interest in US-based synthesis is rational. It removes the border from the risk model, which after August 21 is the fastest-moving risk in the chain. And the chemistry is not the hard part: solid-phase synthesis turns protected amino acids, resin, coupling reagents and cleavage chemistry into API. The hard part is the quality system around it, which is expensive, slow and invisible on a product page. Which is why the claim is made far more often than it is built.

Jay Campbell is the name most associated with this shift, so we read what is documented rather than what is repeated.

He is co-founder of BioLongevity Labs, based in Akron, Ohio. The company's own site states that it "manufactures 99% purity research peptides using solid-phase peptide synthesis in USA-registered facilities that follow Good Manufacturing Practices (GMP)," and its about page says it "manufactures all compounds in U.S.-based GMP-certified facilities." An affiliated domain describes the operation differently, as "a USA-based supplier working with USA-registered, cGMP-aligned synthesis facilities."

Those are not the same claim. Manufacturing in a GMP-certified facility and being a supplier working with cGMP-aligned facilities describe different positions in the chain, and "aligned" is not "certified." We looked for the resolving detail, a named contract manufacturer or a facility registration number, and did not find one. Everything sourcing-related traces back to the company's own copy.

Campbell also argues against the strong version of the domestic-synthesis case himself. On his own site in April 2026 he wrote that FDA restrictions on research-use-only peptides "have absolutely NOTHING to do with public health and EVERYTHING to do with market control," and he promotes BioLongevity with a personal discount code. In the South Florida Sun-Sentinel in August 2026 he said people buying RUO peptides for themselves are "doing research on themselves."

BioLongevity also sells FLGR242 and blends it into a house product. So the operation is a real and early instance of the pattern this article predicts, a vendor-coded analogue sold as a signature product with no pharma originator behind it. What we cannot support on the public record is the reading that it demonstrates US-owned synthesis capacity.

That is not a knock and it is not an endorsement. It is what is checkable, which is the only useful thing to publish here.

This is the part that sellers moving to house compounds have not priced in, and it is the strongest argument against our own thesis, so we will make it as well as we can.

Analogue substitution works in the controlled-substance world because scheduling is molecule-specific. List a molecule, and a molecule one bond away sits outside the list until somebody amends it. That ran for thirty years.

The Food, Drug, and Cosmetic Act does not work that way. The test is intended use. A brand-new peptide nobody has ever seen, marketed for human consumption, is an unapproved new drug on day one. It needs no predecessor, no schedule entry and no patent for FDA to reach it. That is exactly the reasoning in the August 24 letters, where the disclaimer failed and the peptide calculator and the bacteriostatic water became the evidence.

So a novel sequence buys relief from Lilly. It buys close to nothing from FDA. It also buys nothing from WADA, whose 2026 prohibited list retains the S0 catchall covering any unapproved drug or peptide, which means a tested athlete cannot invent their way out either.

What a novel compound changes is the economics of enforcement. FDA has finite letters and finite lawyers. A patent holder with tens of billions in obesity revenue has a very different budget and a very direct motive. Taking the pharma plaintiff out removes the fastest and best-funded adversary and leaves the slower one. That is not safety. It is triage, and vendors will do it because it works on the timescale they care about.

There are four lanes a peptide can legally occupy in the US.

• An approved drug. For weight management, semaglutide as Wegovy, tirzepatide as Zepbound, liraglutide as Saxenda. Retatrutide is not on that list, and FDA's statement page, current as of September 1, 2026, says retatrutide and cagrilintide cannot be used in compounding under federal law.
• Compounded from a listed bulk substance, at a 503A or 503B facility, against a prescription.
• A dietary supplement, through the new dietary ingredient pathway.
• A genuine research reagent, sold to a laboratory that does laboratory work.

The interesting movement in 2026 is in the second lane, and it did not go the way FDA staff wanted.

In April 2026 FDA moved a block of peptides off Category 2, the bulk substances that raise significant safety concerns, into a separate withdrawn-nomination table. The list, current as of April 22, includes BPC-157, KPV, TB-500, MOTS-c, semax, epitalon, emideltide, melanotan II, injectable GHK-Cu, CJC-1295 and ipamorelin acetate. A Federal Register notice on April 16 set a Pharmacy Compounding Advisory Committee meeting for July 23 and 24.

At that meeting the committee voted to recommend six of seven peptides for the 503A bulks list: BPC-157, KPV and TB-500 all 8-6 with one abstention, MOTS-c 7-5-2, then epitalon 7-4-1 and semax 8-5-1. Emideltide, DSIP, was voted down 6-7-1.

Two caveats carry the whole story. FDA's own staff scientists opposed adding every single one, citing absent clinical data, batch-to-batch inconsistency and immunogenicity risk, and the committee voted against them anyway; six of the eight recently appointed committee members operate clinics that offer peptides. And the recommendations are nonbinding. Those peptides remain illegal to compound until notice-and-comment rulemaking finishes, which is measured in years.

The same period produced counter-movement. FDA proposed on April 30, 2026 to permanently exclude semaglutide, tirzepatide and liraglutide from the 503B bulks list, and stated in May that it does not intend to put substances nominated on or after January 7, 2025 into the Category framework at all, closing the on-ramp for anything new. At state level the picture inverts: Louisiana enacted Act 374 on May 22, 2026, effective August 1, barring state boards from stopping prescribers providing peptides from a compliant 503B or 503A facility. A state legislature is building the legal lane while FDA narrows the federal one.

So legal peptides do get big. The mechanism is a 503A bulks list expanded by rulemaking, prescriber protections written state by state, and eventually approvals. Every one of those clocks runs in years. That is why we think this is a 2027 and 2028 story, and why anyone selling "it is about to be legal" as a reason to buy today is selling a timeline they cannot support.

We ran this thesis at people whose job was to break it. Here is what survived, because a forecast that only quotes its own supporting evidence is a press release.

The only quantitative series runs the other way. Chainalysis measured six consecutive quarters of growth, straight through the March 2026 warning letters, the Cincinnati seizures and two federal prison sentences. Every enforcement action in this article happened while the money went up. We are calling an inflection with zero post-August data.

The enforcement is a rounding error on its own numbers. Lilly flagged more than 14,000 listings and referred 200-plus entities, then sued six. That is a touch rate near 0.04%, and Lilly's own executive said out loud that six lawsuits will not solve it.

This market is what the last successful crackdown produced. Compounded semaglutide and tirzepatide were killed by the shortage delistings, affirmed on August 27. The demand did not evaporate. It reappeared with a research-use-only label on it. Every prior squeeze in this category moved the trade downmarket and offshore rather than ending it, and the base rate says displacement, not extinction.

The defendants are the replaceable layer. Six US LLCs, four of them Texas entities. An LLC costs a few hundred dollars and a weekend. The synthesis is in China, and Chainalysis shows precursor suppliers moving into this trade because it pays better and is policed less. Closing a storefront does not touch a reactor.

Adaptation is cheaper than the pivot we predict. The rational response to the August letters is to unbundle the bacteriostatic water, delete the dose calculator, strip the claims out of the copy and keep selling the same molecule. That is a day of web work. Moving to novel chemistry costs a company.

And the hardest one: novelty destroys the product. The entire value proposition of a gray retatrutide vial is that it is the same molecule as Lilly's phase 3 compound at a fraction of the price. Strip out the borrowed clinical credibility and what is left is an unknown powder with no data, no forum consensus and nothing to go viral on. A market built on borrowed credibility has a poor reason to buy compounds that have none.

We do not have a confident answer to that last one. It is the strongest argument that the shift we are describing will be cosmetic, a rename rather than a chemistry change, and the March and August warning letters are evidence it already is.

First, the honest part. The 6 to 12 month number is a judgment call, not a derived figure. Nothing in the research has a 6-to-12-month clock attached to it. The six lawsuits sought no injunctive relief, so ordinary federal timing puts motions to dismiss in early 2027 and merits resolution in 2028. The 503A rulemaking runs in years. If you want the number to come from a model, we do not have one, and we are not going to pretend otherwise.

So here is the narrower claim we will actually defend. The legal architecture that made an open, indexed, credit-card, named-molecule catalog viable is being dismantled, and the open catalog is the part that goes first. Where the demand goes after that is the genuinely open question, and displacement is at least as likely as reduction.

If four of those five break against us by mid-2027, this article was wrong and the RUO market absorbed 2026 the way it absorbed 2024 and 2025.

This is a research reference, so the practical part is about reading vendors. The companion brief, the RUO racket, goes deeper on what the label is selling and why the COAs are theater. This section is the short version, updated for what changed in August.

• A rename is a disclosure. When a catalog turns retatrutide into a code, it has published its own read on its legal exposure. FDA has twice put that behaviour in writing, so the move no longer even buys ambiguity.
• A novel compound with a confident claim is the highest-risk object on the shelf. The confidence cannot be checked by anyone, including the seller, and research on a parent protein is not research on an analogue of it.
• A COA behind a paywall is not a COA. If the certificate only appears after payment, it was never part of the decision.
• Identity before purity. An HPLC number says how clean. Mass spec says what. On a new molecule or a new vendor, identity is the only question that matters first.
• Lot-matched, not catalog-matched. A certificate that does not name the lot in the box describes somebody else's box. The Paradigm Peptides sentencing was, in part, a forged-certificate case.
• No grade is an answer. No trials, no PMIDs, no label, no registry entry, no grade. The blank is the information.

Reptides grades peptides on evidence. When the shelf fills with compounds that have no literature, an evidence-graded reference either says so plainly or turns into an advertisement. Compounds with no data will keep getting no grade. The market map and the enforcement timeline stay updated as this runs, because the dates in this piece will keep moving, and we would rather be checked than believed.

The wild west had a good run. What ends it is not a raid. It is a docket, a warning-letter database and a rulemaking calendar, all of which are public, and all of which you can read yourself.

Read the market map → · The RUO racket · How to read a COA

This article is a market and regulatory analysis published for research reference. It is not legal advice, medical advice, or investment advice, and it is not a recommendation to buy, sell, use or administer any compound. Forward-looking statements in this piece are clearly labelled as the reptides forecast and are not statements of fact. Court cases described here are pending and no defendant has been found liable.