GHK-Cu guide
GHK-Cu is two different situations sharing one name. Put on skin, it is a lawful cosmetic ingredient with a published safety assessment behind it5. Put in a syringe, it is an unapproved substance that no human trial has ever tested21213. FDA treats these as separate questions, and so does this page32.
The short answer
- What it is: a three-amino-acid peptide bound to copper, found naturally in blood plasma and used in skincare as copper tripeptide-1.
- Approved? No, nowhere in the United States, by any route. Copper peptide products do appear in FDA listing directories, but listing is self-registration, not approval1.
- In cosmetics: assessed as safe as used, at concentrations typically under ten parts per million5.
- Injected: FDA has published a specific concern about immune reactions to aggregated or impure peptide, and says human data are limited2.
- Best human evidence: two controlled wound trials from the 1990s that disagree with each other7820.
- For wrinkles: no controlled trial has shown a benefit on any objective measure915.
The void at the centre of the injectable market
There is no controlled human trial of injected GHK-Cu. Not a small one, not an old one, not a bad one. Across the trial registry and the published literature, only two reports exist in which anything containing GHK-Cu was injected into people, and in both cases it was one ingredient in a mixture of six or fourteen21213. In the larger of the two, everyone was also taking finasteride and using minoxidil, so nothing in it can be attributed to any single component1012131415.
FDA says the same thing in its own words, and adds a specific mechanism of concern: injectable GHK-Cu may pose a risk of immunogenicity because of the potential for aggregation and peptide-related impurities2. Worth noticing what that is not. The worry regulators have published is not copper poisoning; it is the immune system reacting to what else is in the vial52.
The compounding status has moved recently and is still moving. The nominations for GHK-Cu were withdrawn in April 2026, then partly reinstated in May for every route except injection, and FDA has said it intends to put the question to its advisory committee before the end of February 20273. An official committee outcome was not identified as of September 20, 2026.
What the controlled skin trials actually found
The strongest human evidence for GHK-Cu is in wound healing, and it points both ways. In venous leg ulcers, a three-arm trial in 86 people compared silver sulfadiazine, a 0.4 per cent copper tripeptide cream and an inert vehicle. Silver sulfadiazine worked. The copper tripeptide cream did not separate from the vehicle7820.
In diabetic foot ulcers, a multicentre vehicle-controlled trial reported the opposite: median closure of 98.5 per cent against 60.8 per cent for the vehicle, healing about three times faster, and infections in 7 per cent against 34 per cent. That is a large result. It is also a result whose sample size, allocation, duration and gel strength are all missing from the published record, and the product was a commercial gel evaluated by authors connected to the company that made it814121620.
Two controlled trials, thirty-odd years old, in different wound types, disagreeing. That is the honest state of it.
The wrinkle claim, examined
Copper peptides are sold overwhelmingly for skin appearance, so it is worth being precise about what has been tested. One randomised trial has looked at GHK-Cu on facial skin: 13 people recovering from laser resurfacing, followed for twelve weeks. Computer analysis and blinded evaluators found no difference in redness, no difference in wrinkles and no difference in skin quality. The only significant finding came from patients rating their own skin on an unblinded questionnaire915.
Other cosmetic studies exist, but they test mixtures. A striae study paired a tripeptide with a hexapeptide around laser sessions and reported no statistics at all. A scalp study combined a scrub with a five-ingredient serum, had no control group, and was co-authored by an employee of the manufacturer1012131415. None of them can tell you what the copper peptide did.
The collagen story that underpins all of this is real, but it lives in cell culture and in rats. Fibroblasts in a dish made more collagen when GHK-Cu was added directly to them, peaking at a nanomolar concentration, and wire-mesh chambers implanted under rat skin accumulated more connective tissue when injected with it1617. Neither is a measurement of human skin, and neither involves anything being applied to a surface.
Does it even get through skin?
Two laboratory studies of human skin samples answer this differently, and the difference is instructive. One kept thinned skin under a continuously replenished copper solution for 48 hours and measured substantial penetration. The other compared intact skin against skin pierced with microneedles, and found that through intact skin almost no peptide or copper crossed at all1819.
The reconciliation is not mysterious. The outer layer of skin is a serious barrier to a copper-bound peptide, and published work shows meaningful passage mainly when that barrier is bypassed or when the exposure is far heavier than any cosmetic use.
Two things that are repeated everywhere and are not established
The first is the age-decline figure: that plasma GHK falls from about 200 ng/mL at 20 to about 80 ng/mL at 60. It appears in almost every article about copper peptides. It traces back to a single sentence in a review that carries no citation, written from a company selling copper peptide skincare, and no primary study measuring GHK in plasma by age could be found[unverified]. Even taking the number at face value, nothing links such a decline to any health outcome.
The second is the cosmetic safety conclusion, which is genuine but narrower than it sounds. The expert panel did conclude that copper tripeptide-1 is safe as used. It reached that conclusion by reading across from related peptides: the human patch and skin-sensitisation testing in the report was performed on a different peptide, and the report contains no repeat-dose or carcinogenicity data for copper tripeptide-1 at all5. A conclusion about a trace ingredient in a face cream is not a conclusion about a milligram in a syringe.
There is also no human pharmacokinetic data of any kind. What is known is that GHK is unstable in human plasma and broken down quickly by enzymes there5.
Sport, and a caveat that catches people out
GHK-Cu is not named on the 2026 WADA Prohibited List, and neither is copper. But the growth-factor section is deliberately written to cover unnamed substances with a similar biological effect, and WADA separately warns that topical skincare can contain prohibited peptides under unfamiliar names, prohibited regardless of route6. WADA has published no ruling either way on GHK-Cu specifically. Anyone subject to testing should treat that as an open question rather than a permission.
What would actually settle this, and a registration that will not
What the field needs is straightforward: give one person two identical wounds, treat one with GHK-Cu and one with the same gel base, blind everyone including the people reading the photographs, and see what happens. A trial registered in 2026 describes exactly that design11.
We are not counting on it. While checking a different compound we looked up everything the same sponsor had registered, and found eight studies sharing one hospital, nearly the same start date and an identical completion date, covering most of the compounds in this library. One of them has the words “Mock Study” in its own title. Another reuses the title and the internal study code of a large trial a different company completed and published years ago. Nobody checks what is submitted to a trial registry.
So the honest position is unchanged from where it has stood for fifty years: an ongoing trial would prove nothing anyway until it reported, and this one should not even be counted on to be running.
Where the rest of the detail lives
The interactive library on the homepage carries the fuller reference entry for GHK-Cu, including the amounts and handling notes it holds for every compound. That is general reference information, the same for every visitor.
References
These are the sources behind this page, listed so you can read them yourself rather than take our word for it. Animal and laboratory findings do not establish that something works or is safe in people. Doses and protocols shown on a study card are that study's research parameters, not recommendations. PepCaddie does not sell peptides and does not choose a dose for anyone; this is educational material and it does not replace a qualified healthcare professional.
Vehicle-controlled trial in venous leg ulcers
1992Human randomised trial
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
Three creams were compared in people with venous leg ulcers: silver sulfadiazine, a copper tripeptide cream, and an inert vehicle. Silver sulfadiazine reduced ulcer size significantly. The copper tripeptide cream did not separate from the vehicle at all. This is the properly controlled negative result in the GHK-Cu literature, and it is the reason a positive finding elsewhere cannot be treated as settled.
Limitation: The treatment period and the exact vehicle composition are not given in the published record, and the trial is more than thirty years old.
Randomised, evaluator-blinded, three-arm trial against an inert vehicleView study: Vehicle-controlled trial in venous leg ulcers
Vehicle-controlled trial in diabetic foot ulcers
1994Human randomised trial
Plantar ulcers treated with a GHK-Cu gel closed by a median of 98.5 per cent against 60.8 per cent for the vehicle, closed about three times faster, and became infected far less often, at 7 per cent against 34 per cent. Everyone was on the same aggressive wound-care protocol, and the authors reported that the advantage depended on starting treatment immediately after the wound was first debrided.
Limitation: The published record does not give the sample size, the allocation between arms, the treatment duration or the gel concentration, so the strength of the result cannot be judged independently. The product tested was a commercial gel, and adjacent published records place the same product with a company whose staff appear among the authors, so an industry evaluation of its own product is likely — although no funding statement exists to confirm it.
Multicentre, randomised, evaluator-blinded, vehicle-controlled trialView study: Vehicle-controlled trial in diabetic foot ulcers
Randomised trial after laser resurfacing
2006Human randomised trial
This is the only randomised trial of GHK-Cu on facial skin that could be found, and every objective measure was null. Computer analysis and blinded evaluators found no difference in how quickly redness settled, and no difference in wrinkles or overall skin quality. Everyone improved, because laser resurfacing works. The single significant result, at p = 0.04, came from patients scoring their own skin on an unblinded questionnaire.
Limitation: Thirteen people split across two arms can only detect a very large effect. The one positive result is self-assessment among several measures with no correction for testing them all, and the comparison arm was a standard regimen rather than a matched vehicle. The concentration used is not stated.
Randomised comparison with and without GHK-Cu, blinded evaluators and image analysisView study: Randomised trial after laser resurfacing
Placebo-controlled hair trial of a GHK combination
2016Human randomised trial
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
Hair count rose by 52.6 in the lower-strength group and 71.5 in the higher-strength group against 9.6 on placebo over six months. Hair length and hair thickness, the two structural measures, did not differ between any of the groups. The authors themselves framed the product as a possible complementary agent rather than a treatment.
Limitation: The product combines GHK with 5-aminolevulinic acid, so nothing here separates the peptide from the other ingredient, and it is GHK rather than the copper complex. The comparison that reached significance against placebo was the lower strength, not the higher one, which is the opposite of what a real dose response looks like.
Three-arm, placebo-controlled comparison of two strengths against placeboView study: Placebo-controlled hair trial of a GHK combination
A registered trial of GHK-Cu alone, treated with caution
2026Registered clinical trial
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
On paper this is the design the field has never had for GHK-Cu as a single agent: each participant receives two standardised wounds, one assigned to GHK-Cu gel and one to an identical vehicle, so the comparison happens within the same person, with everyone including the assessors blinded. It is registered as recruiting from February 2026 with primary completion estimated for February 2027.
Limitation: An ongoing trial establishes nothing about whether a treatment works, and this record carries a further problem. This registration needs more scepticism than a registry entry usually gets. The same sponsor has registered eight studies, all at one hospital, nearly all recording the same start date and all the same completion date, covering most of the compounds in this library. One of them carries the words "Mock Study" in its own title, and another reuses the title and internal study code of a large trial a different company completed and published years earlier. Registry entries come from sponsors and are not checked by the registry, so this record is reported here but is not treated as evidence that a trial is really running.
Phase 2, randomised, quadruple-masked, vehicle-controlled, split-woundView study: A registered trial of GHK-Cu alone, treated with caution
Injected hair cocktail, uncontrolled
2018Human observational study
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
This is one of only two published reports in which anything containing GHK-Cu was injected into people. Hair fall fell on a pull test in 83 per cent of participants and total hair count rose significantly at one year. Seventy-five per cent of participants believed the injections were helping, which the paper reports as a result.
Limitation: This is not a study of GHK-Cu. Copper tripeptide-1 is one of six active ingredients in a proprietary growth-factor mixture, its concentration is not stated, and nothing in the design separates it from the other five. There is no control arm, no blinding, and the authors originated the commercial formulation being tested.
Open-label, single-arm pilot study with no control or placeboView study: Injected hair cocktail, uncontrolled
Large retrospective clinic series
2018Human observational study
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
The largest human exposure to an injected mixture containing copper tripeptide, reported from two clinics over six years. Minor complications occurred in 4.2 per cent of patients, mostly slight pain or bleeding at the injection site, and the authors recorded no treatment-related adverse events. Satisfaction was reported by 96 per cent at six months and 80 per cent at twelve.
Limitation: Everyone also took oral finasteride and used minoxidil twice daily, and the injected solution contained fourteen ingredients of which copper tripeptide was one. A retrospective series with no control arm and satisfaction as its main outcome cannot attribute anything to any single component.
Retrospective clinic case series, no control arm, not blindedView study: Large retrospective clinic series
Open-label scalp study by the manufacturer
2024Human observational study
A two-product scalp regimen improved a flaking score by 61 per cent over fifteen days, with reported increases in hair density and thickness. The study is included here because it is typical of the recent GHK-Cu literature rather than because it settles anything: it is the kind of paper a marketing claim is usually built on.
Limitation: One author is employed by the manufacturer of the tested products, as the paper discloses. There is no control arm and no vehicle, so fifteen days of any scalp routine would be expected to improve flaking, and a reported 25 per cent gain in hair thickness over fifteen days is not plausible as real hair growth. The serum contains five listed ingredients and the concentration of copper tripeptide-1 is not given.
Prospective, open-label, single-centre, no control armView study: Open-label scalp study by the manufacturer
Peptide serum alongside laser treatment
2020Human observational study
Adding a peptide serum around laser sessions was reported to increase improvement, reduce darkening afterwards, and raise satisfaction, and every participant reported feeling better about their stretch marks. The published record contains no effect sizes, no p-values and no statistics of any kind.
Limitation: The serum pairs tripeptide-1 with hexapeptide-12 and the record does not state that the tripeptide is copper-bound, so this may not be GHK-Cu at all. Ten people, no control, and the laser is doing most of the work.
Pilot study with no control or vehicle armView study: Peptide serum alongside laser treatment
The original collagen experiment in cell culture
1988Laboratory or in-vitro studyNot human evidence
Collagen synthesis in cultured fibroblasts rose with GHK-Cu in the dish, starting at around a picomolar concentration and peaking at a nanomolar one, without any change in the number of cells. The paper also notes that the GHK sequence itself occurs within type I collagen. This single experiment is the origin of nearly every collagen claim made for copper peptides.
Limitation: Cells in a dish, bathed directly in the peptide. It says nothing about how much reaches a fibroblast through living skin, and the response peaks and then falls, so a higher concentration is not simply a stronger one. One author has long-standing commercial involvement in copper-peptide skincare.
Cell culture experimentView study: The original collagen experiment in cell culture
Rat wound-chamber study
1993Animal studyNot human evidence
Wire-mesh chambers implanted under the skin of rats were injected with GHK-Cu, and the tissue that grew inside them gained dry weight, protein, DNA, collagen and glycosaminoglycans in proportion to the amount given. This is the most-cited demonstration that GHK-Cu builds connective tissue in a living animal.
Limitation: An artificial chamber under rat skin, receiving the peptide by direct injection. It models neither human skin nor anything applied to the surface of it, which is how people actually use this substance.
Implanted wound chambers receiving direct injectionView study: Rat wound-chamber study
Skin penetration measured under heavy exposure
2010Laboratory or in-vitro studyNot human evidence
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
Copper from a GHK-Cu solution crossed skin samples over two days and left a reservoir behind in the skin itself. This is the study usually cited to show that copper peptides get where they need to go.
Limitation: The skin was thinned before testing and kept under a continuously replenished solution for forty-eight hours, which is far more aggressive than dabbing on a serum. The other laboratory study of the same question, using intact skin, found almost nothing crossed.
Laboratory diffusion-cell study on separated skin layersView study: Skin penetration measured under heavy exposure
Skin penetration measured through intact skin
2015Laboratory or in-vitro studyNot human evidence
Measurable amounts of peptide and copper crossed skin that had first been pierced with microneedles. Through intact skin, in the authors words, almost no peptide or copper permeated at all. The point of the paper was that a delivery device is needed, which is itself the finding.
Limitation: A laboratory model over nine hours, not a person wearing a cream. It does establish that the outer skin barrier is a serious obstacle for this molecule.
Laboratory comparison of intact skin against microneedle-treated skinView study: Skin penetration measured through intact skin
Rat wound study with a vehicle comparison
2003Animal studyNot human evidence
The dose above is the amount used in this study. It is research information, not a recommended dose, and PepCaddie does not convert study doses into human doses.
Poorly supplied wounds in rats closed faster with a copper tripeptide gel than with no treatment at all. Against the gel base on its own, the difference reached significance on only two of the thirteen days measured, which suggests most of the benefit came from the gel rather than from the peptide in it.
Limitation: A rat model, and one where the honest reading of the vehicle comparison undercuts the headline. The product tested was a commercial gel made by a copper-peptide company.
Controlled animal study against vehicle and untreated woundsView study: Rat wound study with a vehicle comparison
Drugs@FDA: FDA-Approved Drugs. U.S. Food and Drug Administration; 2026.
Official regulatory sourceChecked 2026-09-21
Read FDA's approved-drug database at the sourceCertain bulk drug substances for use in compounding that may present significant safety risks. U.S. Food and Drug Administration; 2026.
Official regulatory sourceChecked 2026-09-21
Read FDA's safety statement on injectable GHK-Cu at the sourceBulk drug substances nominated for use in compounding under section 503A. U.S. Food and Drug Administration; 2026.
Official regulatory sourceChecked 2026-09-21
Read FDA's compounding nominations list at the sourceWarning letter: Tailor Made Compounding LLC (MARCS-CMS 594743). U.S. Food and Drug Administration; 2020.
Official regulatory sourceMARCS-CMS 594743Checked 2026-09-21
Read FDA warning letter naming GHK-Cu at the sourceJohnson W Jr, Heldreth B. Safety assessment of tripeptide-1, hexapeptide-12, their metal salts and fatty acyl derivatives, and palmitoyl tetrapeptide-7 as used in cosmetics. Cosmetic Ingredient Review Expert Panel; 2014.
Official regulatory sourceChecked 2026-09-21
Read Cosmetic Ingredient Review safety assessment at the sourceThe 2026 Prohibited List. World Anti-Doping Agency; 2026.
Official regulatory sourceChecked 2026-09-21
Read WADA's 2026 Prohibited List at the sourceBishop JB, Phillips LG, Mustoe TA, et al.. A prospective randomized evaluator-blinded trial of two potential wound healing agents for the treatment of venous stasis ulcers. Journal of Vascular Surgery; 1992.
Human randomised trialPMID 1495150 · DOI 10.1067/mva.1992.37086Checked 2026-09-21
Read Vehicle-controlled trial in venous leg ulcers at the sourceMulder GD, Patt LM, Sanders L, et al.. Enhanced healing of ulcers in patients with diabetes by topical treatment with glycyl-l-histidyl-l-lysine copper. Wound Repair and Regeneration; 1994.
Human randomised trialPMID 17147644 · DOI 10.1046/j.1524-475X.1994.20406.xChecked 2026-09-21
Read Vehicle-controlled trial in diabetic foot ulcers at the sourceMiller TR, Wagner JD, Baack BR, et al.. Effects of topical copper tripeptide complex on CO2 laser-resurfaced skin. Archives of Facial Plastic Surgery; 2006.
Human randomised trialPMID 16847171 · DOI 10.1001/archfaci.8.4.252Checked 2026-09-21
Read Randomised trial after laser resurfacing at the sourceLee WJ, Sim HB, Jang YH, et al.. Efficacy of a complex of 5-aminolevulinic acid and glycyl-histidyl-lysine peptide on hair growth. Annals of Dermatology; 2016.
Human randomised trialPMID 27489425 · DOI 10.5021/ad.2016.28.4.438Checked 2026-09-21
Read Placebo-controlled hair trial of a GHK combination at the sourceA phase 2, randomized, double-blind, vehicle-controlled, split-wound study of topical GHK-Cu gel to accelerate re-epithelialization of standardized acute skin wounds in healthy adults. ClinicalTrials.gov (sponsor: Hudson Biotech); 2026.
Registered clinical trialNCT07437586Checked 2026-09-21
Read A registered trial of GHK-Cu alone, treated with caution at the sourceKapoor R, Shome D. Intradermal injections of a hair growth factor formulation for enhancement of human hair regrowth: safety and efficacy evaluation in a first-in-man pilot clinical study. Journal of Cosmetic and Laser Therapy; 2018.
Human observational studyPMID 29482481 · DOI 10.1080/14764172.2018.1439965Checked 2026-09-21
Read Injected hair cocktail, uncontrolled at the sourceTanaka Y, Aso T, Ono J, et al.. Androgenetic alopecia treatment in Asian men. Journal of Clinical and Aesthetic Dermatology; 2018.
Human observational studyPMID 30057663Checked 2026-09-21
Read Large retrospective clinic series at the sourcePatel MN, Patel N, Merja A, et al.. An assessment of the safety, efficacy, and tolerability of a novel scalp treatment regimen combining a hydroxy acid-based scrub and copper tripeptide serum in the management of seborrheic dermatitis in adults. Cureus; 2024.
Human observational studyPMID 39449909 · DOI 10.7759/cureus.70108Checked 2026-09-21
Read Open-label scalp study by the manufacturer at the sourceMcGuinn KP, Ross NA, Wang JV, et al.. Combination tripeptide/hexapeptide serum with 1540 nm nonablative fractional laser for the treatment of striae distensae: a pilot study. Skinmed; 2020.
Human observational studyPMID 33397562Checked 2026-09-21
Read Peptide serum alongside laser treatment at the sourceMaquart FX, Pickart L, Laurent M, et al.. Stimulation of collagen synthesis in fibroblast cultures by the tripeptide-copper complex glycyl-L-histidyl-L-lysine-Cu2+. FEBS Letters; 1988.
Laboratory or in-vitro studyPMID 3169264 · DOI 10.1016/0014-5793(88)80509-xChecked 2026-09-21
Read The original collagen experiment in cell culture at the sourceMaquart FX, Bellon G, Chaqour B, et al.. In vivo stimulation of connective tissue accumulation by the tripeptide-copper complex glycyl-L-histidyl-L-lysine-Cu2+ in rat experimental wounds. Journal of Clinical Investigation; 1993.
Animal studyPMID 8227353 · DOI 10.1172/JCI116842Checked 2026-09-21
Read Rat wound-chamber study at the sourceHostynek JJ, Dreher F, Maibach HI. Human skin retention and penetration of a copper tripeptide in vitro as function of skin layer towards anti-inflammatory therapy. Inflammation Research; 2010.
Laboratory or in-vitro studyPMID 20703511 · DOI 10.1007/s00011-010-0214-4Checked 2026-09-21
Read Skin penetration measured under heavy exposure at the sourceLi H, Low YS, Chong HP, et al.. Microneedle-mediated delivery of copper peptide through skin. Pharmaceutical Research; 2015.
Laboratory or in-vitro studyPMID 25690343 · DOI 10.1007/s11095-015-1652-zChecked 2026-09-21
Read Skin penetration measured through intact skin at the sourceCanapp SO Jr, Farese JP, Schultz GS, et al.. The effect of topical tripeptide-copper complex on healing of ischemic open wounds. Veterinary Surgery; 2003.
Animal studyPMID 14648529 · DOI 10.1111/j.1532-950x.2003.00515.xChecked 2026-09-21
Read Rat wound study with a vehicle comparison at the source