Last reviewed 8 August 2026
Ask whether research peptides are legal in the United Kingdom and the question is aimed at the wrong object. Nothing in the Human Medicines Regulations 2012 attaches legality to a molecule. The Regulations define a class of thing — a medicinal product — and then regulate what may be done with anything falling inside that class. Whether a given vial of lyophilised peptide falls inside it is decided partly by what the substance is and partly by what its seller has said about it.
That second half is where almost all UK enforcement risk in this market actually sits, and it is the half that is entirely within a supplier’s control. The proposition this article sets out, with the statutory and case-law authority for each step, is narrow and precise: the point at which supply becomes unlawful is normally the attachment of a claim, not the identity of the compound. The important qualification — the one that most pages on this subject omit — is the word normally. For a small number of molecules the identity is sufficient on its own, and no amount of careful wording changes that.
The definition, and why it has two limbs
Regulation 2(1) of the Human Medicines Regulations 2012 defines a “medicinal product” as:
“(a) any substance or combination of substances presented as having properties of preventing or treating disease in human beings; or
(b) any substance or combination of substances that may be used by or administered to human beings with a view to—(i) restoring, correcting or modifying a physiological function by exerting a pharmacological, immunological or metabolic action, or (ii) making a medical diagnosis.”
These are conventionally called the presentational limb and the functional limb, and MHRA is explicit that they are alternatives rather than cumulative conditions: “Medicinal products may well fall under both limbs of the definition but falling under either limb is sufficient to classify a product as a medicinal product.” (MHRA Guidance Note 8, A guide to what is a medicinal product, September 2025, section 4.)
Two definitional points widen the first limb considerably. “Disease” is defined at regulation 8 as including “any injury, ailment or adverse condition, whether of body or mind” — so the limb is not confined to named pathologies. And the Court of Justice held in Upjohn v Farzoo (C-112/89) that the functional limb catches products intended to modify physiological function even where no disease is involved at all, giving contraceptives as the example. A supplier who reasons that a compound is safe to describe because the thing being described is not a disease has misread both limbs.
| Limb one — presentation | Limb two — function | |
|---|---|---|
| Statutory trigger | The substance is “presented as having properties of preventing or treating disease” | The substance “may be used by or administered to human beings with a view to restoring, correcting or modifying a physiological function” by pharmacological, immunological or metabolic action |
| Decided by | What is said, shown, implied, linked to or placed beside the product | What the substance is, and what is scientifically established about it |
| Within the seller’s control | Almost entirely | Not at all |
| Effect of a disclaimer | Persuasive evidence, but not conclusive | None |
Limb one: presentation, and what counts as a claim
The presentational limb is broader than the word “claim” suggests in ordinary use. In van Bennekom (Case 227/82) the Court held that substances which are not “indicated or recommended” as suitable for treating disease may nonetheless constitute substances presented for treating or preventing disease. Presentation is assessed as a whole, not sentence by sentence.
Guidance Note 8, section 6, sets out what MHRA weighs under this limb. The list is worth reading closely, because several items catch material that a technically minded supplier would not think of as promotional at all:
- “all claims made for the product, both explicit and implicit, including any made on websites, linked helplines, testimonials, linked publications, or in social media. Implicit claims may include product names”
- “the context in which the claims are made, and the overall presentation”
- “the promotional literature, including testimonials and any literature issued by the person placing the product on the market or on their behalf”
- “the product form … and the way it is to be used”
- “any particular target of the marketing information/advertising material, for example, population groups with, or particularly vulnerable to, specific diseases or adverse conditions”
Section 5 of the same guidance lists forms of marketing that may suggest a product is properly classified as a medicine. Alongside the obvious entries — references to medical conditions, comparison with licensed medicines, testimonials implying medicinal claims — it includes two that bear directly on any supplier publishing scientific material: “references to medical and / or clinical research and testing”, and “juxtaposing with any examples of the above”.
The second of those is a layout rule dressed as a legal test. Two statements that are each unobjectionable can become a claim by proximity. A description of what a study measured, set immediately beside a price and an add-to-basket control, invites the inference that the study is the reason to buy — and that inference is the claim.
The internet is advertising
Guidance Note 8 forecloses the argument that a website is merely informational: “Information on the internet about a product and its uses is not excluded from the definition of ‘advertisement’… Where a product is sold on or has links to a website which presents that product as a medicine, the website will be used by the MHRA as evidence in the determination process. Similarly, where a customer is directed from a website selling a product to another website for more information about the substances contained in a product and their uses, this may also be used by the MHRA as evidence.”
Outbound links are therefore not a safe harbour. A catalogue page that says nothing itself but points the reader to somewhere that says everything has, on MHRA’s stated approach, put that material into evidence.
Third parties, and words that arrive after the sale
The leading authority is Ter Voort (C-219/91), which concerned herbal teas whose packaging and labelling made no medicinal claim at all; the therapeutic properties were asserted in a brochure issued by a separate foundation. The Court held that a product is “presented for treating or preventing disease” in particular “when it is expressly ‘indicated’ or ‘recommended’ as such, possibly by means of labels, leaflets or oral representation” (paragraph 17), and went further on both attribution and timing:
“The fact that the publication is sent, not by the manufacturer or the seller, but by a third party acting on their behalf or in connection with them cannot rule out any intention on the part of the seller or manufacturer to market the product as a medicinal product.” (paragraph 29)
“The mere fact that the publication is sent to the purchaser only at his request is not capable of rebutting such an intention.” (paragraph 28)
Two consequences follow, and both are routinely missed. Material supplied after the transaction counts. And material issued by someone else counts unless that person is genuinely independent — the operative ruling turns on whether the third party “does not act completely independently”. A supplier who keeps a storefront scrupulously clean and then answers a message with the content the storefront omits has not avoided the limb. They have documented it.
Why a disclaimer is persuasive but not conclusive
The most useful single sentence in this area comes from Delattre (C-369/88), summarised in MHRA’s own case-law appendix. The Court held that a product may be regarded as presented as a medicinal product if its form and packaging render it sufficiently similar to a medicinal product “and, in particular, if on its packing and in the information provided with it reference is made to research by pharmaceutical laboratories, to methods or substances developed by medical practitioners or even to testimonials from medical practitioners commending the qualities of the product”. Then:
“A statement that the product is not medicinal is persuasive evidence which the national court may take into consideration but is not in itself conclusive.”
That is the European analogue of the position United States regulators take under their own intended-use rule, and it settles the status of the “research use only” label in UK law. The label is evidence. It is not a defence, and it cannot outweigh a presentation that points the other way. It also explains why stacking such notices is counter-productive: repetition adds no evidential weight, and a page carrying five disclaimers reads as one whose author understood the exposure and chose to paper over it rather than remove it.
Phrases MHRA has treated as claims
Appendix 1 to Guidance Note 8 lists words and phrases that “have all contributed to a determination by the MHRA that the product they were associated with was a medicinal product”, with the caveat that meaning “has to be considered in context” and that the list is not exhaustive. A representative selection, with MHRA’s own stated reading:
| Phrase | What MHRA states it may suggest or imply |
|---|---|
| “Alleviates” | “In context, may suggest a claim to treat disease by reducing, ameliorating or correcting disease or an adverse condition.” |
| “Boosts” | “In context, claim may tend to suggest that the product may be administered with a view to modifying physiological function and having a significant effect.” |
| “Help/help with…” | “In context, may be a claim to treat, provide relief from, and cure symptoms of disease or an adverse condition.” |
| “Is said to help with…” | “In context, may be an implied claim to efficacy in relation to disease or adverse condition.” |
| “Medical research…” | “An implied claim to efficacy as a medicine.” |
| “Clinical Trials Evidence” | “Implied claim to (medicinal) efficacy in relation to disease or an adverse condition.” |
| “Clinically proven” | “In context, a claim to work directly to treat, prevent or cure disease or an adverse condition.” |
The last three entries deserve emphasis, because they are the trap for a supplier who believes that citing literature is the conservative option. Under MHRA’s stated approach, a bare gesture at “medical research” is an implied efficacy claim. What distinguishes a lawful reference from an unlawful one is not the presence of a citation but its grammar. “A 2011 study reported increased fibroblast outgrowth in cultured rat tendon explants” is a statement about a study: it names the model, reports what was measured, and asserts nothing about the substance in a human being. “Backed by research” is a claim about the product with the study removed. The first is a factual report; the second is the thing Appendix 1 lists.
Limb two: function, which no wording switches off
The functional limb is assessed on the substance, and a supplier’s copy is close to irrelevant to it. HLH Warenvertriebs and Orthica (C-211/03) requires national authorities to proceed case by case, “taking account of all the characteristics of the product, in particular its composition, its pharmacological properties, to the extent to which they can be established in the present state of scientific knowledge, the manner in which it is used, the extent of its distribution, its familiarity to consumers and the risks which its use may entail”.
The threshold is real rather than theoretical. Hecht-Pharma (C-140/07) held that a product cannot be a medicinal product by function where, having regard to its composition and if used as intended, it is “incapable of appreciably restoring, correcting or modifying physiological functions”, and that the Directive does not apply where medicinal status has not been scientifically established, “even if the possibility cannot be ruled out”. Commission v Germany (C-319/05) put it more shortly: the functional definition covers products “whose pharmacological properties have been scientifically observed”. And in the joined cases of Markus D. (C-358/13) and G. (C-181/14) the Court held that substances producing effects that merely modify physiological functions, without beneficial effects on human health, fall outside the definition altogether.
Read together, these cases mean that for a compound with a thin human evidence base, the functional limb is genuinely uncertain — and uncertain in the supplier’s favour, since the burden is on the authority to establish pharmacological properties rather than on the supplier to disprove them. That is a real legal point, and it is also the point at which honesty requires the argument to be turned around, because the same reasoning is decisive against a small number of compounds.
Guidance Note 8’s list of limb-two factors includes “the presence of essentially similar licensed, registered or exempt medicines on the UK market”. Where a molecule is the active ingredient of an authorised medicine, its pharmacological properties are not merely established, they are established in a dossier the licensing authority itself assessed. Nothing a supplier writes or declines to write alters that. The determination machinery reflects it: Guidance Note 8, Appendix 4 records that MHRA may determine a product to be a medicinal product without following the statutory review-panel procedure where, among other circumstances, “the product is a copy of, or is identical in all material respects to… an existing licensed or registered medicine”.
Within this catalogue that reasoning reaches exactly two entries, and both are handled under a fixed form of words and a separate sign-off rather than by ordinary editorial judgement. For human chorionic gonadotrophin and for Melanotan I: the same molecule is the active ingredient in a UK-authorised medicine — Pregnyl and Ovitrelle in the first case, Scenesse in the second. NovoVita’s material is not that medicine, is not manufactured to medicinal standard, and is not supplied for human or veterinary use. Four further molecules that correspond to authorised medicines or to controlled drugs are, for the same reason, not published in the compound library at all pending a regulatory decision.
The offences, and the machinery behind them
Regulation 46 is the operative prohibition. It provides that a person may not sell or supply, or offer to sell or supply, an unauthorised medicinal product; may not supply a medicinal product otherwise than in accordance with a UK marketing authorisation or equivalent; and, at regulation 46(3), may not possess an unauthorised medicinal product knowing or having reasonable cause to believe it is intended to be supplied to another person in the United Kingdom or the EEA. Regulations 46(4) and (5) extend the same prohibition to manufacturing, assembling or procuring those acts with that knowledge.
Regulation 47 supplies the offence and the penalty: a person who breaches regulation 46 is guilty of an offence, liable on summary conviction to a fine not exceeding the statutory maximum, and on conviction on indictment to a fine, to imprisonment not exceeding two years, or to both.
Advertising is prohibited separately. Regulation 279 provides that a person may not publish an advertisement in Great Britain for a medicinal product unless a marketing authorisation, certificate of registration or traditional herbal registration is in force for it — and, per Guidance Note 8, “advertisement” is defined broadly at regulation 7 and is not read as excluding website content. The structural consequence is worth stating plainly: a claim does two things at once. It brings the substance inside the definition of a medicinal product, and the same words then constitute an advertisement for an unauthorised medicine.
Classification itself runs through Part 9 of the Regulations. MHRA issues a provisional determination with reasons; the company may make written or oral representations to an independent Review Panel; a final determination follows, and regulation 163 gives power to require a company to stop marketing a product. Breach of such a notice can itself be a criminal offence. Regulation 165 allows MHRA to bypass that procedure where there is an identifiable risk to public health, where the product is identical in all material respects to an existing licensed medicine, or where “advertising material for a product makes clear medical claims… and the company responsible has not voluntarily complied with a MHRA request to remove or reword the material within a set timescale”.
That third route is the one that matters commercially. The realistic first contact with the regulator is not a raid; it is a request to reword a page, and the consequence of refusing is the loss of the procedural protections that would otherwise apply.
Where the Misuse of Drugs Act sits on top
Medicines law is not the only regime in play, and the second one is easy to overlook because it attaches to substances that are not obviously “drugs” in the colloquial sense. Paragraph 4 of Part II of Schedule 4 to the Misuse of Drugs Regulations 2001 lists, alongside the anabolic and androgenic steroids, the following: chorionic gonadotrophin (HCG), non-human chorionic gonadotrophin, somatotropin, somatrem and somatropin. These are Class C controlled drugs under the Misuse of Drugs Act 1971.
Two features of the Schedule 4 Part II regime are commonly misread.
The possession exemption is narrow, and it is about possession only. Regulation 4(3) of the 2001 Regulations disapplies the section 5(1) possession offence for a Schedule 4 Part II drug in the form of a medicinal product. It says nothing about production or supply, which remain offences under section 4 of the 1971 Act unless licensed — and unlawful supply of a Class C drug carries a maximum of fourteen years on indictment, materially more than the two years available under the Human Medicines Regulations. There is also an unresolved question on the face of the drafting: the exemption is framed by reference to a drug “in the form of a medicinal product”, and material presented as a laboratory reagent is, by the supplier’s own account, not that. Reading the exemption as covering research-labelled material is an argument, not a settled position, and it is not one this article endorses.
The import and export exemption is narrower still. Regulation 4(2) is confined to importation or exportation “which is carried out in person for administration to that person” of a Schedule 4 Part II drug. A consignment moving by post or courier is outside its terms.
This is the reason somatropin is the one line in the catalogue blocked from leaving the United Kingdom, enforced at cart validation and again at checkout rather than stated as a policy and left to good faith. A published restriction that nothing enforces is worse than none.
What is left when nothing is claimed
If neither limb is engaged, the material is not a medicinal product and the Human Medicines Regulations do not reach it. It is a chemical, and it falls to the regimes that govern chemicals: UK REACH and the classification, labelling and packaging rules, hazard communication and safety data sheets where applicable, and the General Product Safety Regulations 2005, which Guidance Note 8 identifies as the residual framework and which are enforced by Trading Standards rather than by MHRA.
Guidance Note 8 draws the boundary in terms that are helpful to a supplier who intends to stay on the right side of it. Section 7 records that MHRA “only classifies finished products and not individual substances and ingredients”, and that “a product will not be classified as a medicine solely on the basis that it may be unsafe for human use. A product must be intended for, or be capable of performing, a medicinal function before it can be classified as such.”
But section 11 removes any comfort that the position, once reached, is durable:
“The Agency reserves the right to change its view in the event of any information or evidence which has a bearing on the status of the product, including the way in which it is packaged, promoted or presented, or if there is a change in scientific knowledge or the law.”
Non-medicinal status is therefore not a licence, a registration or an achievement. It is a description of how a product is currently being presented, and it lapses the moment the presentation changes. It is maintained by conduct, continuously, or not at all.
The same analysis, applied to this catalogue
The constraints below are NovoVita’s published editorial rules. They are set out here not as a compliance advertisement but because a worked example is more useful than an abstraction, and because each one can be traced to a specific limb and a specific authority rather than to caution in general.
| Constraint | Limb | Authority |
|---|---|---|
| Product copy states identity, presentation, storage and research-use status only — no indications, no outcomes, no populations | One | GN8 §6 limb-one factors; regulation 2(1)(a) |
| No heading may name a condition, symptom, body system or population | One | GN8 §6: implicit claims include product names; a heading is presentation marked up for machines |
Structured data is restricted to a permitted set of types; Drug, MedicalWebPage and MedicalIndication are excluded by rule | One | Presentation is not confined to prose; a machine-readable claim is portable and indexed |
| Literature is reported as what a named study measured in a named model, never as a property of the substance | One | GN8 Appendix 1: “Medical research…”, “Clinical Trials Evidence” |
| No literature statement is placed within one screen of a price, a basket control or a discount banner | One | GN8 §5: “juxtaposing with any examples of the above” |
| The reconstitution tool computes volume from concentration; presets fill vial strength only, and no target amount, frequency or duration is offered | One | GN8 §6: “the way it is to be used”; Delattre on information provided with the product |
| Purity is published as a composition statement with its boundary stated explicitly | One | A statement of what is in a vial is a different legal category from a statement of what it does |
| The research-use notice is rendered once by the template and never repeated | One | Delattre: such a statement is persuasive but not conclusive, so repetition adds nothing |
| Somatropin is blocked from despatch outside the United Kingdom at cart and checkout | Separate regime | MDR 2001, Sch 4 Pt II; MDA 1971 s.3 and s.4 |
| The two authorised-medicine matches carry one fixed construction and require named sign-off | Two | GN8 limb-two factor: essentially similar licensed medicines on the UK market |
The rule that generates most of the others is a grammatical one, and it is worth stating in isolation because it survives translation to any catalogue: a sentence may be a statement about a study, about a rule, or about a molecule — it may not be a statement about what the product will do. Entries such as BPC-157, TB-500 and GHK-Cu are written to that constraint: sequence, molecular formula, mass, CAS number, solubility, storage behaviour and analytical method are properties of the substance and are stated flatly; everything drawn from the literature is attributed to a named publication and a named model.
An equivalent line governs the arithmetic. A reconstitution calculation converts a mass and a volume into a concentration, and that is a fact about a solution. Selecting a per-administration amount is a judgement about a subject. NovoVita’s internal formulation of the boundary — the volume is ours to calculate, the target amount is not — is why competitor-style “typical” presets were declined rather than adopted, notwithstanding that they are common in this market and that they would perform.
What none of this does is touch limb two. Every constraint in the table above reduces exposure under the presentational limb only. For a molecule whose pharmacological properties are established, the functional limb is engaged by chemistry and the copy is beside the point. Any supplier claiming that careful wording makes their whole catalogue lawful has either not read the second limb or is hoping the reader has not.
Enforcement as it actually looks
MHRA’s Criminal Enforcement Unit is active in this specific market. On 29 May 2026 the agency reported the seizure of approximately 12,000 doses of unlicensed medicines, including retatrutide and tirzepatide together with peptide products, packaging materials and substances believed to be used in illicit manufacture, following a raid at a property near Northampton on 28 May 2026. Two men were arrested on suspicion of offences under the Human Medicines Regulations 2012 — not under any bespoke peptide legislation, because none exists. The regulations quoted earlier in this article are the ones under which the market is actually policed.
The position in the United States runs on the same logic through different machinery. Under the Food and Drug Administration’s intended-use regulation for drugs, 21 CFR 201.128, intended use is determined by the objective intent of the persons legally responsible for labelling, evidenced by labelling claims, advertising, oral or written statements, and the circumstances surrounding distribution — including where a seller has knowledge of facts giving notice that an article is used for a purpose for which it is neither labelled nor advertised. The FDA has issued warning letters to retailers of research-labelled peptide products on that basis, among them letters listed in its public warning-letter database to USApeptide.com (26 February 2025), and to Prime Sciences and Gram Peptides (both 31 March 2026). The mechanism differs from Delattre; the conclusion is the same. A label stating that a product is not for human use is one piece of evidence about intent, and it is outweighed by conduct pointing the other way.
What this analysis does not settle
Three limits should be stated, since a page whose value is precision cannot afford to overclaim about itself.
The functional status of most of these compounds is genuinely open. Hecht-Pharma requires pharmacological properties to be established in the present state of scientific knowledge, and for many peptides in circulation they are not. That protects a supplier under limb two today. It equally means the evidence base is thin, and it does not become thicker because it is legally convenient. A 2026 narrative review of twelve approved and unapproved peptides used in musculoskeletal contexts — Mendias CL, Awan TM. Sports Medicine, 12 April 2026 (Epub), PMID 41966639 — reported that although many of these unapproved peptides had shown favourable tissue-repair and metabolic outcomes in animal models, rigorous human safety data were scarce, and identified that gap as carrying potential for harm. A supplier is entitled to rely on the legal consequence of an incomplete evidence base only if it is willing to state the incompleteness.
Determinations are product-specific and are not precedent. MHRA classifies on a case-by-case basis, and section 3 of Guidance Note 8 records that “no single factor or combination of factors will necessarily be conclusive… But in relation to particular products, a single factor or combination of factors may be more important than others, and may even be conclusive.” That a comparable product was not classified as medicinal tells a supplier very little about their own.
This is analysis, not legal advice. It is NovoVita’s reading of published sources, offered because the reasoning behind the site’s own constraints ought to be inspectable rather than asserted. It reflects the law as published at the date shown and cannot substitute for advice on a specific product. MHRA itself attaches the same caveat to Guidance Note 8, which “should not be taken as a complete or definitive statement of the law”. Anyone who needs certainty about a particular product can request a formal opinion through MHRA’s borderline advice route.
Sources
- Human Medicines Regulations 2012, regulation 2 (definition); regulation 46; regulation 47; regulation 279.
- MHRA Guidance Note 8, A guide to what is a medicinal product, published September 2025 — sections 3, 4, 5, 6, 7 and 11, Appendix 1 (words and phrases), Appendix 3 (case law) and Appendix 4 (statutory determination procedure).
- Case C-219/91 Ter Voort; Case Case 227/82 van Bennekom; Case C-369/88 Delattre; Case C-112/89 Upjohn v Farzoo; Case C-211/03 HLH Warenvertriebs and Orthica; Case C-319/05 Commission v Germany; Case C-140/07 Hecht-Pharma; Joined Cases C-358/13 and C-181/14 Markus D. and G.
- Misuse of Drugs Regulations 2001, regulation 4 and Schedule 4; Misuse of Drugs Act 1971, sections 3, 4 and 5.
- MHRA news, “Two arrested during the MHRA’s largest ever seizure of unlicensed weight loss medicines”, 29 May 2026.
- 21 CFR 201.128 (meaning of intended uses); FDA warning-letter database.