Peptides run from supplements you can buy today to the biggest drugs in metabolic medicine. What they are, how the categories differ, and where to start.
Peptides are short chains of amino acids, and that single chemistry fact stretches across supplements, cosmetics, prescription medicines and research chemicals. Sorting which category you are looking at is the skill that makes everything else readable.
“Peptide” is a chemical class, not a safety or legality stamp.
Clean, prescription, and research products follow different rules.
Start with bucket and evidence, not influencer vocabulary.
A peptide is a short chain of amino acids. That is the whole definition, and it is a statement about chemistry - not about what a product does, whether it is legal, or whether it works.
This matters more than it sounds. The word “peptide” is doing an enormous amount of marketing work right now, and it covers a collagen powder in a supermarket, a prescription medicine that has been through clinical trials, and a vial sold online with a label saying it is not for human consumption. All three are peptides. Almost nothing else about them is alike.
This page is the map. It explains what a peptide actually is, how peptides act in the body, and the three-bucket model we use across this site to keep those very different things apart. It also says, plainly, where the evidence runs out.
What a peptide actually is
Amino acids are small molecules. Link two of them together and the bond that forms is called a peptide bond. Link a handful and you have a peptide; link a great many and the same kind of chain is called a protein. Everything else - the shape it folds into, whether it survives digestion, what it binds to - follows from which amino acids are in the chain and in what order.
Two real examples show the range.
GHK, the peptide at the heart of copper-peptide skincare, is three amino acids: glycine, histidine and lysine. PubChem lists its molecular formula as C14H24N6O4 and its molecular weight as 340.38 g/mol, with a computed XLogP of −4.4 and a topological polar surface area of 176 Ų. Those numbers describe a very small, very water-loving molecule.
BPC-157, the most talked-about research peptide, is usually described as a synthetic peptide of fifteen amino acids. It is roughly four times the length of GHK and behaves nothing like it.
Where peptides stop and proteins start
Here is the first honest gap on this page: the peptide/protein boundary is a naming convention, not a law of chemistry. You will see “under 50 amino acids” quoted confidently across the internet. We could not source that cutoff to an authority we would be willing to cite, so we are not going to state it as fact.
What we can show is how arbitrary the line is in practice. Thymosin beta-4 is a 43-amino-acid molecule and is generally called a protein. The compound sold as TB-500 is Ac-LKKTETQ - a seven-amino-acid, N-acetylated synthetic fragment corresponding to residues 17 to 23 of thymosin beta-4 - and is called a peptide. Our source here is a secondary encyclopedia entry, not a primary paper.
That is not a trivia point. It is the single most useful thing on this page, because vendors routinely borrow the parent protein’s research to sell the fragment. A seven-residue fragment is not a functional stand-in for a 43-residue protein, and the studies do not transfer automatically.
How peptides work in the body
Lead with the honest version: mechanism is the easiest thing to write about and the least informative. A plausible pathway in a dish tells you almost nothing about whether a product changes anything in a person.
Then there is the delivery problem, which decides more arguments than mechanism does. Bos 2000 - Bos JD, Meinardi MM. The 500 Dalton rule for the skin penetration of chemical compounds and drugs. Exp Dermatol. 2000;9(3):165-9. (opens PubMed in a new tab) proposed a rule of thumb: a compound must be under 500 daltons to be absorbed through skin, because larger molecules cannot pass the corneal layer. The support offered is three observations about compounds we already have - that common contact allergens, common topical dermatological drugs, and all known transdermal drugs are under 500 daltons. That is a heuristic drawn from what exists, not a measured physical limit. GHK at 340 Da clears the bar - but its XLogP of −4.4 and TPSA of 176 Ų make it extremely hydrophilic and highly polar, which independently disfavours passive penetration of a lipid barrier. Whether GHK-Cu reaches the living dermis at cosmetic concentrations in a real person appears never to have been settled empirically; we could not locate a human penetration study.
The three buckets
This is the model the rest of the site runs on. Open each card.
Collagen & skinReal human trials behind them, and you can buy them today. Sold as supplements or cosmetics.Collagen peptides · Copper peptides (GHK-Cu) · Topical signal peptides
How you'd get it
Buy it in a shop or online today. No prescription, no grey market.
Where it stands legally
Sold legally as supplements (collagen) or cosmetics (topical peptides). Neither category is FDA-approved before sale - that is normal and it is not a scandal, but it does mean the burden of evidence sits with the manufacturer.
Prescription / GLP-1The most studied peptides there are. Approved medicines, large human trials, prescription only.Semaglutide · Tirzepatide
How you'd get it
Through a licensed clinician who assesses whether it is appropriate for you.
Where it stands legally
FDA-approved medicines that have been through clinical trials. Legal, well-studied, and prescription-only for good reason - they have real effects and real side effects.
Research peptidesThe frontier. Striking early findings, mostly in animals, and not approved for human use.BPC-157 · TB-500 · CJC-1295 · Ipamorelin
How you'd get it
Sold online by "research chemical" vendors, labelled not for human use.
Where it stands legally
Not approved for human use in the US. The "research use only" label is the legal basis on which vendors sell at all - it is not a wink, it is the thing that keeps the sale lawful. Purity and contents are not verified by any regulator.
Open one to see how it's sold, and where it stands legally.
The buckets are a reader’s shorthand. The legal spine underneath them is narrower and sharper: intended use plus route of administration.
Federal law defines a drug partly by intended use - articles intended for the diagnosis, cure, mitigation, treatment or prevention of disease, or intended to affect the structure or any function of the body (FD&C Act §201(g)(1)). A cosmetic is defined by cleansing, beautifying, promoting attractiveness or altering appearance (§201(i)). FDA states that the law does not recognize any category called “cosmeceuticals” - the term has no meaning under the Act.
Route does the rest of the work. A dietary supplement must be, in the statute’s words, intended for ingestion (§201(ff)(2)(A)(i)). That one clause defeats most grey-market “it’s just a supplement” framing without needing any peptide-specific ruling: an injectable peptide cannot lawfully be marketed as a dietary supplement, whatever the ingredient.
Bucket one: the clean peptides
US legal status - Dietary supplement
Regulated as a dietary supplement in the US. Supplements are not FDA-approved; manufacturers are responsible for their own safety and labelling, and the FDA acts only after a problem surfaces.
These are the ones you can buy today, and the two that matter are ingested collagen and topical cosmetic peptides.
Multiple randomised trials, though with limitations. Multiple double-blind randomized trials and two meta-analyses in over 1,100 participants each, for oral collagen and skin. Effects are statistically real, small in size, and inconsistent across endpoints; much of the flagship literature carries manufacturer-affiliated co-authors.
US legal status - FDA-approved · prescription only
An FDA-approved prescription medicine. It has been through clinical trials for its approved use, and legally requires a prescription from a licensed clinician.
The GLP-1 medicines - semaglutide, tirzepatide - are peptides too. They are the reason the word is in the culture at all. They have been through clinical trials, they are FDA-approved, and they require a prescription because they have real effects and real side effects.
The interesting part for this site is the shadow market beside them. As of May 31, 2026, FDA reported 990 adverse event reports associated with compounded semaglutide and more than 730 associated with compounded tirzepatide, while noting that state-licensed pharmacies which are not outsourcing facilities are not required to report, so the true count is likely higher. FDA has also described dosing errors from patients measuring and self-administering incorrect doses, fraudulent products labeled with pharmacies that do not exist, counterfeit Ozempic, and products arriving warm. FDA’s baseline advice is quotable: compounded drugs should only be used in patients whose medical needs cannot be met by an FDA-approved drug.
Bucket three: the research peptides
US legal status - Research use only · not for human use
Sold for laboratory research use only. Not approved by the FDA for human use, and cannot be lawfully marketed as a supplement or a medicine. Products sold this way are not made to pharmaceutical quality standards, and their contents are not verified by anyone.
BPC-157, TB-500, CJC-1295, ipamorelin. None is FDA-approved for any indication in the United States. They are sold as research chemicals, labeled not for human use, and bought by people who intend to use them on themselves.
Small or uncontrolled human studies. Too little to settle the question either way. For BPC-157, the largest research-bucket literature: rodent and in vitro work, with the review literature stating efficacy is yet to be confirmed in humans. The human evidence is too small and too underpowered to answer the question: five studies in roughly 80 people, of which the single randomized placebo-controlled trial was null and exists only as a conference abstract.
Three further points people rarely hear. First, under §503A a compounder may only use a bulk drug substance that complies with an applicable USP or NF monograph, or - if no monograph exists - is a component of an FDA-approved drug, or - if neither applies - appears on FDA’s 503A bulks list; a substance meeting none of those conditions may not lawfully be used in compounding. Second, FDA’s Pharmacy Compounding Advisory Committee is scheduled to meet on July 23–24, 2026 to discuss BPC-157, KPV, TB-500 and MOTs-C, and Emideltide/DSIP, Semax and Epitalon, for possible inclusion on that list - the evaluated use for BPC-157 being ulcerative colitis, not injury recovery. This page is accurate as of July 15, 2026 and that section is expected to age. Third, if you compete: BPC-157 is named in the WADA 2026 Prohibited List under S0, and CJC-1295, ipamorelin and thymosin-ß4 derivatives including TB-500 are named under S2 - all prohibited at all times.
If you are reconstituting anything, our dosage calculator shows the arithmetic and, more importantly, shows where the arithmetic stops helping.
A short word on safety
Safety is not a property of the word “peptide”. It is a property of a specific molecule, at a specific dose, by a specific route, in a specific person - and the honest ranking is roughly the buckets themselves: an approved medicine has a known side-effect profile, a supplement has a long history of ingestion, and a research chemical has neither a human safety database nor a verified content list.
Our full treatment, including the regulatory picture and FDA’s own red-flag list for consumers, is in Are peptides safe?.
What we don’t know
This is the part of the page the rest of the internet skips.
Where the peptide/protein boundary actually sits. Widely quoted, and we could not source it to an authority. Treat any number as a convention.
Whether topical peptides reach the tissue they are supposed to act on. GHK is small enough by the 500 Dalton heuristic and strongly hydrophilic against it. No human penetration study resolving this was located.
What dose of a topical peptide does anything. Products are sold across a wide concentration range. We are not aware of human dose-response data for GHK-Cu.
Whether collagen’s effect is specific to collagen. The flagship trials use carbohydrate placebos rather than an equivalent-protein control, so we could not find a trial testing whether ordinary protein would do the same thing.
Why the 2026 meta-analysis found elasticity inconsistent when elasticity was the older literature’s headline result. Better bias-handling, newer null trials, or different inclusion criteria - we can’t tell, and it determines how confidently anyone should state the skin claim.
What is actually in a research-peptide vial. We could not locate published analytical work testing purity or contamination of BPC-157 or TB-500 products specifically. Extrapolating from GLP-1 test-buy data would be an inference, so we won’t make it.
Whether independent groups can replicate the BPC-157 findings. The efficacy literature is dominated by one research group. That fact is descriptive; it does not prove the findings wrong, and the literature alone cannot distinguish a funding gap from a replication problem.
None of these gaps is a reason to panic. Each is a reason to distrust anyone who speaks about peptides with total confidence - including us, on any sentence where we have not shown you the source.
References
Bibliographic detail is fetched from PubMed, not written by us - so a citation here cannot drift from the paper it names. Study design comes from PubMed's own tags rather than our judgement.
