THE COMPARISON
Two Routes, Two Registers, One Misleading Word
Mechanism, evidence maturity and regulatory status set side by side — and the distinction that marketing most reliably erases.
The short version
The three compounds on this desk get grouped together because they all involve peptides and all get described as being about skin. That grouping hides almost everything worth knowing.
GHK-Cu is a copper-carrying molecule rubbed onto skin. It is a legal cosmetic ingredient, it has small human trials behind it, and its main technical problem is simply getting through the skin's outer barrier.
Melanotan II is a hormone copy injected into the body. It is approved nowhere, it acts on receptors far beyond the skin, and the recent literature about it is largely case reports of harm.
GLOW is a mixture of three research peptides in one injection, including the same copper molecule that is legal in a cream. Nothing has been tested in that combination.
So the useful question is never is this peptide good or bad. It is: which route, which register, and how mature is the evidence for that specific use.
Side by side
| GHK-Cu | Melanotan II | GLOW (research blend) | |
|---|---|---|---|
| What it is | Copper(II) chelate of a natural collagen-derived tripeptide | Cyclic, protease-resistant analogue of alpha-MSH | Co-formulated mixture: GHK-Cu + BPC-157 + TB-500 |
| Primary target | Dermal fibroblasts, matrix enzymes, copper-dependent lysyl oxidase | MC1R-MC5R melanocortin receptors, non-selectively | Three targets at once: matrix, VEGFR2 vasculature, G-actin |
| Route in practice | Topical, in cosmetic formulation | Subcutaneous injection | Subcutaneous injection |
| Effect on colour | Chemical and local; copper is a tyrosinase cofactor | Signalled and systemic; drives eumelanin synthesis without UV | Incidental; the blend is marketed on radiance, not pigment |
| Strongest human evidence | Placebo-controlled hair-count trial in 45 men [3]; small topical dermatology trials [1][4] | Double-blind crossover in 10 men, erectile endpoint [12] | None for the blend; constituent reviews only [13][14] |
| Documented serious harms | None attributed in the peer-reviewed record for topical use | Renal infarction [10]; rhabdomyolysis, priapism, melanoma in case reports | Untested; inherits constituent risk profiles |
| Regulatory register | Legal cosmetic ingredient as Copper Tripeptide-1; no approved drug indication by any route | No approval by any regulator, anywhere; explicit regulator warnings | Not approved; BPC-157 flagged by FDA as not eligible for compounding pending evaluation |
| Anti-doping | Not itemised; catch-all category may apply | Prohibited under the non-approved-substances category | Prohibited — thymosin beta-4 is listed [13] |
The row that does the most work is the last three. Two compounds sit outside every approval framework, and the third is legal only inside one narrow register — and only there.
Two registers, not one ladder
The most common analytical error in this category is treating regulatory status as a single ladder — approved at the top, cosmetic somewhere in the middle, research chemical at the bottom. It is not a ladder. Cosmetics and medicines are separate registers with separate rules, and a compound can occupy one, both or neither.
An INCI name is a naming convention, not a verdict. Copper Tripeptide-1 on an ingredient list means the ingredient is permitted in cosmetic products in that jurisdiction and that the product is sold on appearance claims. In the United States, cosmetic ingredients are not pre-approved the way drugs are. So the presence of that name establishes permitted cosmetic use and a marketed safety history, and it establishes nothing at all about efficacy as a medicine, about any other route of administration, or about what a systemic dose does. GHK-Cu has no FDA- or EMA-approved therapeutic indication by any route.
The reverse inference fails just as badly. A compound being unapproved as a drug does not make it unlawful as a cosmetic ingredient, and the harms documented for a systemic melanocortin agonist say nothing about a topical copper complex. These are different molecules doing different things by different routes.
A third failure mode is inheritance by association. Melanotan II is regularly defended by pointing at its relatives: afamelanotide is approved for a rare light-sensitivity disorder, and bremelanotide, derived from Melanotan II itself, is approved for a sexual-desire indication [11]. Both approvals are real. Neither transfers. Approval attaches to a specific molecule with a specific trial package for a specific indication — not to a chemical family, and not to a mechanism.
The cleanest statement of the boundary is this: regulatory status attaches to a route and a claim, never to a molecule's reputation. The same GHK-Cu that is a legal cosmetic ingredient in a serum becomes an unapproved systemic agent the moment it is reconstituted for injection, and the GLOW blend is precisely that crossing made routine.
Evidence maturity, honestly ranked
Ranking these three by how much is actually known produces an order that has almost nothing to do with how confidently each is marketed.
GHK-Cu, topical, is the most mature — and 'most mature' here still means small trials. There is a randomised, placebo-controlled six-month hair-count study in forty-five men, though of a combination formulation rather than the pure peptide [3]; small controlled dermatology studies reporting improvements in laxity, clarity, firmness, density and wrinkle depth [4]; a quantified human skin-penetration study establishing dermal delivery [5]; and a deep preclinical mechanism running from fibroblast culture through to transcriptome-wide analysis [2][6][7]. The honest caveats are that much of the foundational literature comes from a single research group, and that the delivery problem the 2025 review centres on is not solved [1].
Melanotan II has the best-characterised mechanism and the worst clinical package. The receptor cascade is textbook, the central appetite effect has been isolated in a well-controlled animal paradigm [9], and the human erectile effect was demonstrated in a double-blind crossover design [12]. But no Phase II or Phase III trial has been completed, so the recent literature consists largely of case reports — renal infarction [10], oral-mucosal pigmentation [8], and the melanocytic and priapism reports described on its own page.
GLOW is the least evidenced, by a wide margin. There is no controlled study of the combination at all. Its most data-poor component has three small human pilot studies behind it and is described in its own 2025 review as investigational [14], and the review that names all three constituents together concludes that rigorous human safety data are scarce with potential for serious harm [13].
Marketing confidence, in this category, runs in almost exactly the opposite order to evidence.
Where the marketing blurs
Three specific blurs recur often enough to be worth naming.
The first is the word peptide itself. It describes a size class, not a pharmacology. A cosmetic tripeptide, an injected hormone analogue and a supplier-mixed blend share the word and share nothing else — not evidence, not route, not register, not risk.
The second is natural. GHK genuinely occurs in human collagen and genuinely declines with age [4]. That is a real and interesting fact, and it is not an argument. Endogenous origin says nothing about the safety of a synthetic version delivered by an unstudied route at a concentration the body never produces locally.
The third is glow and radiance, which travel freely across the boundary. They are cosmetic words attached, in the case of the GLOW blend, to a systemic injectable containing two unapproved research chemicals and a WADA-listed fragment [13]. The vocabulary belongs to one register; the contents belong to another.
Reading this category well mostly means refusing those three substitutions and asking the boring questions instead: which molecule, by which route, tested in whom, and measured how.