The short answer: four states, four different evidence bars
Approved means a regulator examined evidence that a specific product is safe and effective for a specific use, and agreed. In the United States that is a new drug application or a biologics license application, and the standard written into the statute is substantial evidence from adequate and well-controlled investigations. It is product-specific and indication-specific, and it produces a label.
Cleared is a device word and only a device word. Most medical devices reach the American market through premarket notification, in which the manufacturer shows the FDA that its device is substantially equivalent to a device already legally on sale. Nobody has to prove the new device helps anyone. The claim being evaluated is a comparison to an existing product, not a clinical outcome.
Authorised is the ambiguous one. In American usage it most often means an emergency use authorisation, issued during a declared emergency under a standard that the product may be effective, and revocable the moment the agency changes its mind. In European and British usage a marketing authorisation is the full approval, the direct equivalent of an FDA approval. The same word therefore marks the weakest and the strongest state depending on which regulator issued it.
Unapproved is the fourth state, and it is where most of the peptides discussed in enthusiast forums live. No regulator anywhere has evaluated the product for any use in humans. That is not the same as banned and not the same as proven dangerous, but it does mean nobody outside the seller has looked at what is in the vial.
What approval actually requires
An approval application is a dossier, not a permission slip. It has to characterise what the molecule is and how it is made, show that manufacturing is reproducible at commercial scale, present the animal toxicology, and then present human trials that measure a prespecified endpoint against a control. For most indications regulators expect more than one adequate and well-controlled study, or one very persuasive one with supporting evidence.
The output is a label, and the label is the boundary of what was approved. Semaglutide is approved as a weekly injection for type 2 diabetes, as a daily tablet for the same condition, and at a higher weekly dose for chronic weight management. Those are separate approvals with separate trials behind them, because a regulator approving a molecule for one purpose has said nothing at all about a second purpose. When the same molecule later earned a cardiovascular risk-reduction indication, it took a dedicated outcomes trial in more than seventeen thousand people to get there.
Approval also carries obligations that survive the decision: adverse event reporting, post-marketing surveillance, sometimes a formal risk management programme, and manufacturing inspections. Withdrawal is possible and does happen. This is the part that never applies to a product nobody approved, which is why the safety record of an unapproved peptide is not merely thin but structurally uncollectable.
Cleared: a comparison, not a demonstration
Premarket notification exists because most devices are iterations of things already sold. A manufacturer identifies a predicate device, shows the new one has the same intended use and raises no new questions of safety or effectiveness, and receives a clearance. Clinical data appears in a minority of these submissions, and equivalence chains can run back through many generations of predicate to a device that itself never had to prove anything.
Genuinely high-risk devices take a different route: premarket approval, which does require valid scientific evidence of safety and effectiveness and is the only device pathway that earns the word approved. A third route, De Novo classification, exists for novel low-to-moderate-risk devices with no predicate, and produces a marketing authorisation rather than a clearance or an approval. Three device outcomes, three different words, and only one of them is approval.
This matters for peptides mostly at the point of sale. A microneedling pen or a light therapy device sold alongside a peptide serum may be legitimately 510(k)-cleared, and the clearance says nothing whatever about the serum. The FDA has repeatedly told companies that describing a cleared device as approved is misbranding, which tells you how often the substitution is made and how commercially useful it is.
Authorised: the word that means two opposite things
An emergency use authorisation is a wartime instrument. It requires a declared emergency, no adequate approved alternative, and a finding that the known and potential benefits outweigh the known and potential risks on the totality of available evidence. The effectiveness standard is that the product may be effective, deliberately lower than the substantial evidence required for approval, because the alternative during an emergency is nothing at all.
Authorisations are also reversible in a way approvals rarely are. The hydroxychloroquine authorisation issued early in the COVID-19 pandemic was revoked within months once trial data arrived. The authorisation for bamlanivimab used alone was revoked once resistant variants made it unreliable. Meanwhile the same emergency produced the clearest demonstration of the difference: a COVID-19 vaccine authorised in December 2020 and fully approved in August 2021 under a biologics license, same product, different evidentiary state.
In Europe the word runs the other way. The European Medicines Agency recommends and the European Commission grants a marketing authorisation, the equivalent of an FDA approval, with the same dossier requirements and the same indication-bound label. The British regulator issues authorisations on the same model. So a product described as authorised without naming the regulator is telling you almost nothing.
Unapproved: the default state of most peptides sold online
A product is a drug, in American law, if it is intended to diagnose, treat, cure, mitigate or prevent disease, or to affect the structure or function of the body. Intent is inferred from how the thing is marketed, which is why the label on a vial does not decide its legal category. A vial marked for research use only that is sold with recovery, fat loss or longevity claims is, on that reasoning, an unapproved new drug being marketed with claims nobody has evaluated.
Research use only labelling comes from the world of laboratory reagents and in vitro diagnostics. It signals that a material has not been validated for any clinical use. It is a disclaimer of applicability, not a licence to sell into human use, and it does not create a lawful category for a compound intended to be injected by the buyer.
The dietary supplement route is closed to most of these compounds too. A substance first authorised for investigation as a new drug, with substantial public clinical investigation behind it, is generally excluded from the dietary supplement definition. That exclusion catches many peptides with any clinical history at all, and the FDA has issued warning letters on that basis.
The consequence is not merely legal. Unapproved means there is no reviewed manufacturing standard, no required identity or purity testing, no adverse event system that captures what happens, and no label at all. The absence of a regulatory verdict is often marketed as the absence of a problem. It is instead the absence of information.
Compounded material sits outside the ladder entirely
Compounded preparations are made by pharmacies operating under section 503A or by registered outsourcing facilities under section 503B. Neither is FDA-approved, and neither claims to be. Compounding exists for the patient who cannot use the approved product as supplied, because of an allergy to an excipient, a needed strength that is not manufactured, or a dosage form the label does not offer.
The law explicitly bars compounding a copy of a commercially available approved drug, with a narrow exception when the drug is on the FDA shortage list. That exception is what produced the compounded semaglutide market: while the approved products were in shortage, copies were lawful. When the agency declared the shortage resolved in early 2025, the exception closed, with short wind-down windows for 503A pharmacies and 503B facilities running into that spring.
The important point for a reader comparing options is that a compounded peptide is not a generic. A generic is an approved product that had to demonstrate bioequivalence to the reference drug. A compounded preparation demonstrates nothing to anyone; its quality is a function of the facility that made it, and the two facility categories differ substantially in the standards they must meet.
Where four named peptides actually sit
Semaglutide is approved. It cleared the full pathway in the United States and Europe, first for type 2 diabetes and later at a higher dose for chronic weight management, with a 68-week randomised trial showing roughly fifteen percent mean body weight reduction against about two percent on placebo, and a cardiovascular outcomes trial showing about a fifth fewer major adverse cardiovascular events. Those trials are why the word approved attaches to it.
Cenegermin is also approved, and it is a useful contrast because the indication is narrow and the evidence base is small. It is a recombinant human nerve growth factor eye drop for neurotrophic keratitis, a rare corneal disease. It was authorised in the European Union in 2017 and approved by the FDA in 2018 as a biologic, on randomised vehicle-controlled trials in which roughly two thirds to three quarters of treated eyes achieved complete corneal healing against roughly a third on vehicle. Small trials, rare disease, real approval.
Teduglutide is a glucagon-like peptide-2 analogue approved in 2012 on both sides of the Atlantic for short bowel syndrome, with paediatric use added later. Its pivotal work measured something concrete: the volume of intravenous parenteral support patients still needed after treatment. An approved peptide indication can be extremely specific and still be the whole of what was proven.
BPC-157 is unapproved everywhere. There is no published randomised placebo-controlled human efficacy trial, no regulator has evaluated it, and in 2023 the FDA placed it in the category of bulk drug substances that raise significant safety risks, which removed it from lawful pharmacy compounding in the United States. It is the most discussed peptide in the recovery market and it occupies the weakest regulatory state that exists.
Approved is not one thing either
Accelerated approval is a real approval granted on a surrogate endpoint reasonably likely to predict clinical benefit, with confirmatory trials required afterwards. It is how many oncology and rare disease drugs reach patients years earlier than they otherwise would. It is also how a product can be approved and later withdrawn: the amyloid-lowering Alzheimer's antibody approved on that basis in 2021 was contentious from the day of the decision and was eventually discontinued by its sponsor.
Off-label use is the other place where the word stretches. Prescribing an approved drug for a condition outside its label is lawful and common, and in some fields it is standard care. But the approval covers the product, not that use. When a peptide is described as approved for a purpose its label does not name, what is being borrowed is the credibility of an evaluation that was never performed for that purpose.
Phrases in marketing copy that carry no regulatory weight
FDA-registered is the most common. Facilities that manufacture drugs register with the agency; registration is an administrative listing, and the FDA states plainly that it does not denote approval or endorsement of the products made there. Made in an FDA-inspected facility is the same claim with a different verb: an inspection is an event, not a verdict on a specific product.
Pharmaceutical grade has no legal definition in the United States. GMP-manufactured describes a process standard and says nothing about whether the compound works, or whether a given batch met it. Third-party tested is only as good as the assay and the laboratory, and certificates of analysis are among the easiest documents in this market to forge.
Clinically studied is the subtlest of them, because it can be literally true and still carry no weight. A compound with one open-label case series in eleven people has been clinically studied. So has a compound with three phase 3 trials. The phrase compresses those into one claim, which is exactly why it is used. The honest questions are how many people, randomised against what, measuring what, for how long, and published where.
What the distinction changes for a reader
The four states map onto four different kinds of uncertainty. With an approved product, the uncertainty is whether the average trial result applies to you, and there is a label and a surveillance system to consult. With a cleared device, the uncertainty includes whether the thing does anything at all, because nobody was required to show that. With an emergency authorisation, the evidence base is provisional by design.
With unapproved material, the uncertainty is unbounded, and it is not only about efficacy. Identity, purity, sterility, the counterion, the actual peptide content of the powder and the endotoxin load are all unverified, and none of them are visible to the buyer. Reasonable people can decide that a particular unapproved compound is worth their attention, but that decision should be made with the category named accurately rather than obscured by a borrowed word.
The practical habit worth forming is simple. When a claim uses one of these words, ask which regulator, on which pathway, for which indication, and in which year. Any product with a real approval can answer all four instantly, because the answers are printed on the label.