A “Research Use Only” Label Is Not a Legal Status: What FDA and Lilly Established in 2026
FDA set aside an RUO disclaimer and ruled on the rest of the website instead. Lilly then asked payment processors and carriers to cut off sellers. Together they define how this sector is now assessed.
In March 2026 the FDA issued a warning letter that set aside a seller's ["Research Use Only"](https://condorresearch.com/research/what-research-use-only-means/) labelling in a single sentence and determined the products' status from the rest of the website instead. In August, Eli Lilly filed six lawsuits against US sellers of retatrutide and publicly asked payment processors, e-commerce platforms and logistics carriers to cut off the trade — explicitly naming sellers who "falsely claim their drugs are for 'research-use only' when they are actually intended for human use." Read together, these establish the operating principle for this sector: a disclaimer does not determine legal status; the totality of the presentation does. That principle is not new law, but 2026 is the year it was enforced at scale.
Condor Research supplies characterised reference materials for laboratory research use only. This article is regulatory analysis, not legal advice. For the retatrutide clinical data, see our separate piece on the TRIUMPH-2 and TRIUMPH-3 results; this one is about the enforcement environment, and contains no purchasing or dosing information.
The doctrine: intended use is inferred, not declared
US drug law does not define a product by what its label disclaims. Under section 201(g)(1) of the Federal Food, Drug and Cosmetic Act, something is a drug if it is intended for use in the diagnosis, cure, mitigation, treatment or prevention of disease, or intended to affect the structure or function of the body. And intended use, under 21 CFR 201.128, is established by the objective evidence of what the seller means the product for — which includes advertising, website copy, and the surrounding circumstances of the sale.
That is why a disclaimer cannot do the work people expect it to do. It is one piece of evidence about intent, sitting alongside every other piece. If the other pieces point the other way, it loses.
Case study: the Gram Peptides letter
FDA’s warning letter to Gram Peptides, issued 31 March 2026, is the cleanest illustration available, and every operator in this sector should read the original.
FDA reviewed the company’s website between January and March 2026. The products carried the expected labelling: “Research Use Only”, and “not intended for human consumption, medical use, or veterinary use”. The agency’s response is one sentence, and it is the whole doctrine:
“Despite statements on your product labeling marketing your products for ‘Research Use Only,’ … evidence obtained from your website establishes that your products are intended to be drugs for human use.”
What supplied that evidence was the product copy. Phrases FDA quoted include language about appetite suppression, insulin sensitivity, thermogenesis and fat oxidation; about “reducing body weight, improving glucose handling, and enhancing lipid metabolism”; and, on a second product, “Weight Management – Associated with significant decreases in body weight in both animal and human studies.”
Note what is doing the damage. Not a dosing table. Not an instruction. Descriptions of physiological effect, written in a register aimed at a person rather than an experiment. Under 201(g)(1), “intended to affect the structure or function of the body” is a complete definition of a drug on its own — the disease half is not required.
The finding that generalises furthest: adjacency
One paragraph of that letter has implications well beyond the peptides themselves. FDA determined that the company’s bacteriostatic water was also a drug, and the reasoning was purely structural:
“Your firm offers ‘Bacteriostatic Water for Injection’ for sale alongside peptide products, which are drugs intended for injection and require reconstitution… The sale of these products together demonstrates that you intend for your ‘Bacteriostatic Water for Injection’ to be used in combination for injection.”
No claim was made about the water. No copy described its effects. Its intended use was inferred from what it was sold next to. That is worth stating plainly, because it means catalogue architecture is a compliance surface: a product’s regulatory character can be determined by the products adjacent to it, by what a bundle implies, and by what a related-products module recommends.
The commercial escalation: attacking the infrastructure
On 12 August 2026 Eli Lilly moved on the same problem from the private side.
The company filed six lawsuits in Texas and California federal courts, naming compounding pharmacies, medical spas and online sellers. It stated it had referred more than 200 individuals and entities to FDA, the Department of Justice, state attorneys general, law enforcement and professional licensing boards, and had reported more than 14,000 websites, advertisements, social posts and product listings across more than 100 countries.
The strategically significant part was not the litigation. It was the request that followed. Lilly called on:
- social media and e-commerce platforms, to block the conduct before it reaches consumers;
- credit card companies, payment processors, shipping and logistics companies — “and other entities that sellers use to conduct their illegal businesses” — to cut off the infrastructure;
- regulators, customs agencies and law enforcement, to prioritise enforcement across borders.
That is an attack on the rails rather than the trains. And it matters more than the lawsuits, because infrastructure providers do not adjudicate individual cases. A payment processor confronted with reputational and chargeback risk in a category does not distinguish carefully between a compliant reagent supplier and a non-compliant one; it makes a category decision, and appeals are slow.
Lilly’s release also quotes the FDA position that so-called research-use-only products “are of unknown quality and may be harmful” to consumers’ health. Whatever one thinks of that as a blanket characterisation of laboratory reagents, it is now the framing a payments risk committee will encounter first.
What follows, practically
Not moralising, and not a claim to be above the sector. Four observations that follow directly from the documents.
The disclaimer is the weakest part of a compliance position, not the strongest. It is necessary. It establishes nothing on its own. Anyone whose compliance rests primarily on a footer has misunderstood 21 CFR 201.128.
The copy is the compliance position. What determines regulatory character is whether the whole presentation is consistent with laboratory use: how effects are described, whose perspective they are described from, whether the framing is a model system or a person. “In rodent models, researchers observed…” and “supports recovery” are not stylistic variants. They are different legal claims.
Adjacency is a decision, not a default. Reconstitution supplies, bundles, “frequently bought together” modules and related-product blocks all generate the same kind of evidence FDA used for bacteriostatic water. Which products sit beside which is an editorial choice with regulatory consequences, and it should be made deliberately rather than inherited from a plugin’s defaults.
The enforcement risk in 2026 is commercial before it is legal. The realistic first consequence for most operators — the one that shapes the grey market more than any court does — is not a letter from a regulator. It is a payment processor, a platform, or a carrier making a category decision. That is a faster mechanism with no due process, and Lilly has explicitly asked for it.
The honest position
There is a legitimate research reagent trade. Laboratories need characterised materials, those materials are not medicines, and supplying them with identity and purity documentation, no therapeutic claims and no consumer framing is a lawful activity in most of the EU. That is a real distinction, and it is not the one FDA was attacking in March 2026.
What FDA attacked was a seller describing appetite suppression and weight reduction on a product page and then adding a disclaimer. The distinction between that and a reagent business is not the disclaimer. It is everything else on the page.
The useful test is the one FDA actually applies: if a reader who had never seen the disclaimer looked at the site, what would they conclude the products are for? That is the question the regulator asks, and it is the only version of the question that matters.
Related reading
- FDA's March 2026 warning letter to Gram Peptides set aside “Research Use Only” labelling in a single sentence and determined the products' status from the seller's own website copy.
- Under 21 CFR 201.128, intended use is established by objective evidence — advertising, website copy and the circumstances of sale — not by a disclaimer.
- Descriptions of physiological effect were enough: “intended to affect the structure or function of the body” is a complete definition of a drug, with no disease claim required.
- FDA also ruled the seller's bacteriostatic water a drug purely because it was sold alongside injectables requiring reconstitution, which makes catalogue adjacency a compliance surface.
- On 12 August 2026 Lilly filed six lawsuits and asked payment processors, platforms and carriers to cut off the trade — a category-level risk that moves faster than any court.
Does a “Research Use Only” label make a sale legal?
No. It is one piece of evidence about intended use. FDA set aside exactly that labelling in March 2026 and determined the products' status from the rest of the seller's website.
What is 21 CFR 201.128?
The US regulation defining “intended use”. It establishes intent from objective evidence, including labelling, advertising and the circumstances surrounding the sale.
Why did FDA call bacteriostatic water a drug?
Because it was sold alongside injectable products that require reconstitution. Intended use was inferred from adjacency, with no claim made about the water itself.
Does this apply in Europe?
Not directly — FDA's jurisdiction is US commerce. EU law reaches a similar place through the “medicinal product by presentation” limb of Directive 2001/83/EC, and payment processors apply US-derived risk frameworks globally.