Do Copper Peptides Grow Hair? Only Minoxidil Has Human Trials

Minoxidil has randomized trials in hundreds of people with hair counts attached, and neither copper peptide, GHK-Cu or AHK-Cu, has a single published human hair-growth trial.¹²³⁴ That gap is the whole comparison. It is not close.

Minoxidil's own record is also weaker than its reputation. Roughly three in five people who use it see nothing.²

What human evidence does minoxidil have?

Minoxidil carries FDA-approved status for hair regrowth and a body of randomized trials behind it. Its OTC label states the indication as a "Hair regrowth treatment for men," for use "to regrow hair on the top of the scalp (vertex only)."¹

Dias and colleagues collect the trial figures in a 2026 review in Frontiers in Pharmacology. For men, it describes a 48-week study "in which 393 men with vertex hair loss were randomised to receive 5% minoxidil solution," reporting that "the 5% minoxidil group showed a mean increase of 18.6 non-vellus hairs/cm2 from baseline," set against "12.7 hairs/cm2 in the 2% minoxidil group and 3.9 hairs/cm2 in the placebo group."² For a 16-week foam trial in "352 men with AGA," it reports that "hair counts in the target area increased by 13.4% in the treatment group versus 3.4% in the placebo group."²

Women have their own trials. The review describes "a pivotal RCT of 381 women with FPHL treated with topical minoxidil 5%" with "Changes in non-vellus hair counts from baseline were 24.5, 20.7 and 9.4 in a 1 cm2 target area," and a second trial of "404 women with FPHL" in which foam "resulted in regrowth of 10.9 hairs/cm2 and 9.1 hairs/cm2 more than a vehicle foam" at 12 and 24 weeks.²

Every one of those figures reaches this page through the review's account of the original trials. Its reference list did not render on the page read here, so the primaries behind each number stayed unverified at source.² One smaller trial was read directly: a 2016 study of 5% foam in men reported target-area hair counts rising from 162.85 at baseline to 194.58 at week 24 by camera count, at p=.019, and from 54.92 to 69.91 by eye count at p=.002.⁵ That study had no control group. It enrolled 17 men with 12 completing, so it carries far less weight than the randomized work.⁵

Why does minoxidil fail in most people who use it?

Minoxidil needs an enzyme to work. Many scalps do not have enough of it. That Frontiers in Pharmacology review describes the drug as "a prodrug that requires bioactivation by sulfotransferase (SULT) enzymes," naming "specifically SULT1A1, in the outer root sheath of HFs."²

Without that step the drug does nothing at all. "This enzymatic conversion is critical; without sulphation, minoxidil remains pharmacologically inactive," and "higher follicular SULT1A1 activity is associated with a better response to minoxidil therapy," while "non-responders often lack measurable enzymatic activity."²

How often it fails is the number most people never hear. Twice the same review states that "approximately 60% of patients are considered 'non-responders'" and that "approximately 60% of users do not achieve visible improvement."² A drug can have the best evidence in its category and still do nothing for the majority, and both halves of that sentence are true here.

Minoxidil's mechanism is also not settled, even after decades of use. The review grants that "its precise mechanism of action in promoting hair growth remains incompletely understood."²

Does a stronger minoxidil work better?

Stronger minoxidil did not outperform 5% in the trial that tested them head to head. Singh and colleagues, in a 2022 review in Clinical and Experimental Dermatology, report a 36-week study of 90 men comparing 5%, 10% and placebo, where the mean ratio change in total hair count at the vertex was 0.47 plus or minus 0.26 on 5%, against 0.05 plus or minus 0.13 on 10% and 0.01 plus or minus 0.05 on placebo.⁶

The 10% arm performed closer to placebo than to 5%, and that runs against the assumption that concentration drives response. The 2022 review reports that trial without a p value. Its own account of another study carries an internal inconsistency: the text says patients were treated with 5% while the table lists 15%.⁶ A review that contradicts itself on a dose is weaker evidence than one that does not, and that discrepancy belongs next to the finding.

What happens when you stop?

Hair regrown on minoxidil is lost when you stop, and the label says so without hedging: "If you stop using ROGAINE Extra Strength Solution, you will lose your newly regrown hair in 3 to 4 months."¹ The carton adds that "continued use is necessary to increase and keep your hair regrowth, or hair loss will begin again."¹

Duration of benefit has a ceiling in the trial record too. The label states that "hair regrowth has not been shown to last longer than 48 weeks in large clinical trials with continuous treatment."¹

Two separate claims sit inside that sentence and they get conflated constantly. One is about reversal, meaning the regrowth goes away when treatment stops. The other is about the evidence horizon, meaning the trials simply did not run long enough to say what happens in year three. A drug can keep working indefinitely and still have a 48-week evidence ceiling, and a reader cannot tell which situation applies from the label alone.¹

The effect on the comparison with a copper peptide is direct. Minoxidil's dependence on continued use is documented. A copper peptide has no documented duration of benefit to lose, because no human hair trial has established one.³⁴

What evidence does AHK-Cu have?

Evidence for AHK-Cu comes to one published study, and that study used tissue and cells instead of people. Pyo and colleagues, in a 2007 paper in Archives of Pharmacal Research, tested L-alanyl-L-histidyl-L-lysine-Cu2+ on human hair follicles ex vivo and on cultured human dermal papilla cells in vitro.³ No human subjects were involved.

Within that setting the measured effects were real. AHK-Cu "stimulated the elongation of human hair follicles ex vivo" across concentrations from 10^-12 to 10^-9 M, and it increased dermal papilla cell proliferation.³ Markers moved in the direction the authors expected, with the Bcl-2/Bax ratio rising and cleaved caspase-3 and PARP falling at 10^-9 M.³

Isolated follicles in a dish and follicles in a scalp are different systems. Dishes have no blood supply, no hormones arriving from elsewhere, no immune activity and no skin barrier for a topical product to cross. Elongation in that setting establishes that the molecule does something to the tissue. It does not establish that a serum applied to a head grows anything.

Did the AHK-Cu apoptosis result reach significance?

No. Apoptosis of dermal papilla cells fell at 10^-9 M, and the authors report that "this decrease was not statistically significant."³

That detail matters because inhibiting apoptosis is half of the mechanism the paper proposes. The authors' conclusion is that AHK-Cu may promote hair follicle growth by stimulating dermal papilla proliferation and inhibiting apoptosis, and the apoptosis half of it rests on a result that missed.³ Pages citing this study as evidence that AHK-Cu prevents follicle death are citing a non-significant finding.

Supporting markers did move, and they are a weaker class of evidence than the endpoint itself. Bcl-2 to Bax ratio rose, and cleaved caspase-3 and PARP fell at the same concentration, all of which point the same direction as reduced apoptosis.³ Marker changes describe a mechanism in progress; the apoptosis measurement is the thing the mechanism was supposed to produce, and it missed.

Concentration is the other detail that rarely travels with the citation. The effects appeared from 10^-12 to 10^-9 M, picomolar to nanomolar, somewhere between a trillionth and a billionth of a mole per liter.³ Those are vanishingly small amounts, and a bathing concentration around an isolated follicle is not the same quantity as a concentration applied to intact scalp, where the barrier, the sebum and the dilution all intervene before anything reaches a dermal papilla.

What evidence does GHK-Cu have for hair?

GHK-Cu has no published human hair-growth trial that could be located through any permitted route for this guide. Searches returned product pages and affiliate articles, and no peer-reviewed study of hair outcomes in people surfaced.

Absence of a trial is not proof that GHK-Cu does nothing. It does mean that anybody describing GHK-Cu as having hair evidence is describing something other than a trial in humans, and the distinction between a mechanism and an outcome is the one that matters to somebody choosing what to put on their scalp.

GHK-Cu does have a real research literature, and almost none of it concerns hair. The 2018 review summarizes work on skin regeneration, wound healing, nerve and blood vessel growth, antioxidant and anti-inflammatory activity, lung protection and proteasome activity, drawing largely on in vitro and animal studies with some clinical data from cosmetic trials.⁴ A molecule can be well studied in one domain and unstudied in another.

Treating the first as evidence for the second is the move this whole topic turns on.

Why do so many pages say GHK-Cu grows hair?

Pages saying GHK-Cu grows hair trace the claim to a review that asserts it and cites no study. Pickart and Margolina's 2018 paper in International Journal of Molecular Sciences lists among GHK-Cu's established actions the ability to "Increase hair growth and thickness, enlarge hair follicle size," and states that GHK-Cu is "widely used in skin and hair products."⁴

No human study of hair is described anywhere in that paper.⁴ The growth claim appears as a bullet point in the introduction, with no supporting study attached to it in the text.⁴ The authors are affiliated with R&D Skin Biology, a commercial developer of GHK-Cu products. That affiliation does not make the claim false. It does mean the most-quoted source for it has a stake in it.⁴

Bullet points in a review's introduction become citations in product pages, and product pages become what a search engine returns. Nothing in that chain adds a measurement.

Are GHK-Cu and AHK-Cu the same thing?

Different peptides, and not the same thing: the sequences differ and so does the evidence behind them. GHK-Cu is glycyl-histidyl-lysine bound to copper, and AHK-Cu is L-alanyl-L-histidyl-L-lysine bound to copper.³⁴ One amino acid differs at the first position.

For hair specifically the two have separate records. The better one belongs to AHK-Cu. Pyo and colleagues tested AHK-Cu in 2007, not GHK-Cu.³ Citing that study as GHK-Cu evidence substitutes one peptide for another, and treating a GHK-Cu review's hair bullet as applying to AHK-Cu does the same thing in reverse.

One amino acid is not a trivial difference in a three-residue peptide. Glycine and alanine differ by a single methyl group, and in a tripeptide that group is a third of the variable structure, and it changes how the molecule sits in a copper coordination complex and how it presents to a receptor. Short peptides are not interchangeable the way a longer protein tolerates a conservative substitution.

Commercially the two are also sold differently, and that is where most of the confusion enters. Copper peptide is used as a category label across products containing either one, so a shopper comparing two serums described as copper peptides may be comparing different molecules with different evidence, and the ingredient list is the only place the distinction appears.³⁴

Can a copper peptide be compared with minoxidil at all?

Not on evidence, because the two sit at different stages. Minoxidil has randomized placebo-controlled trials in several hundred people each with hair counts as the endpoint.² AHK-Cu has one laboratory study.³ GHK-Cu has an uncited claim in a review.⁴

Set side by side, the three sit at different stages of evidence entirely.

Comparing them produces a one-sided result, and a fair framing of that answer includes minoxidil's 60% failure rate.² Somebody whose scalp lacks SULT1A1 activity gets nothing from the drug with the best evidence, and that person's interest in alternatives is reasonable. What follows from it is that no alternative has been shown to work, not that the alternatives are equivalent.

Non-response is also partly predictable, and that changes what a non-responder should do next. Follicular sulfotransferase activity can be assayed, and the 2026 review cites work finding that "patients with high follicular sulfotransferase activity were significantly more likely to respond to treatment."² Knowing the enzyme status answers whether topical minoxidil was ever going to work, a different question from whether something else will.

Other minoxidil routes exist as well. The same review covers oral and sublingual formulations, and reports a 9-month trial in which low-dose oral minoxidil raised mean hair fiber diameter by 4 μm and density by 13 per square centimeter, with the differences from topical 2% not reaching statistical significance.² Those are prescription routes carrying their own risk profile. They sit inside the same drug class.

What would a real comparison require?

A real comparison would require a head-to-head trial in people, with hair counts, a control arm and a defined duration. Nothing of that kind has been published for either copper peptide against minoxidil or against vehicle.

Several pieces would need to exist first. A topical formulation with a stated peptide concentration and demonstrated delivery through skin, an endpoint measured the way the minoxidil trials measured it, and enough participants to detect a difference the size the minoxidil trials detected, under 20 hairs per square centimeter in the best case.²

What has not been measured?

Four gaps run through this topic.

Scalp penetration for either copper peptide has no published figure, so a serum's label concentration has no established relationship to what reaches a follicle. Regrowth outcomes for GHK-Cu in people have no trial at all.⁴ AHK-Cu's one study stopped at tissue and cells, and its apoptosis arm missed significance.³ Minoxidil's own mechanism remains "incompletely understood" by the account of a 2026 review, four decades into its use.²

Two further gaps belong to the copper peptides specifically. Neither has a published dose-response curve on any hair endpoint in people, so no concentration can be called adequate or inadequate. Neither has a head-to-head trial against anything, including vehicle, so even a positive result in a single-arm study would not separate the peptide from the carrier it arrived in.³⁴

What exists instead is a mechanism, a plausible one, and a commercial literature built on top of it.

Common questions

Do copper peptides regrow hair?

Which has better hair evidence, GHK-Cu or AHK-Cu?

Is minoxidil proven to grow hair?

How often does minoxidil fail?

Why does minoxidil not work for everybody?

Does a 10% minoxidil work better than 5%?

What happens if you stop minoxidil?

How long does minoxidil keep working?

Are GHK-Cu and AHK-Cu the same peptide?

Where does the claim that GHK-Cu grows hair come from?

Who wrote the main GHK-Cu review?

Did the AHK-Cu study show it stops follicles dying?

Does an ex vivo follicle study predict what a serum does on a scalp?

Has anybody compared copper peptides with minoxidil directly?

Is minoxidil's mechanism understood?

What would it take to show a copper peptide grows hair?

Sources

Sources

1. Men's ROGAINE Extra Strength Solution, minoxidil 5% w/v. US over-the-counter Drug Facts labeling. Read at source from the FDA label PDF on accessdata.fda.gov. Note that this is an OTC Drug Facts label and carries no clinical studies section, so it supports the approved indication, the time-to-result statements and the discontinuation statements, and not the hair count figures. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10
2. Dias FR, Yong SS, FitzGerald H, Sinclair RD, Bhoyrul B. "Expanding the therapeutic landscape of minoxidil for androgenetic alopecia: topical, oral and sublingual formulations." Frontiers in Pharmacology 2026;16:1718208. DOI 10.3389/fphar.2025.1718208, published 21 January 2026. Read at source. The minoxidil trial figures quoted from it are this review's account of the original trials, and its reference list did not render on the page read here, so each primary stayed unverified at source. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17 ↩18 ↩19 ↩20 ↩21 ↩22 ↩23 ↩24 ↩25 ↩26 ↩27 ↩28
3. Pyo HK, Yoo HG, Won CH, Lee SH, Kang YJ, Eun HC, Cho KH, Kim KH. "The effect of tripeptide-copper complex on human hair growth in vitro." Archives of Pharmacal Research 2007;30(7):834-839. DOI 10.1007/bf02978833. PMID 17703734. Abstract read via an institutional repository record, not at the publisher. The repository lists the author names in a different order from the citation above, so author order should be checked against the journal before republication. The full text was not read. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17 ↩18 ↩19 ↩20 ↩21 ↩22 ↩23 ↩24 ↩25
4. Pickart L, Margolina A. "Regenerative and Protective Actions of the GHK-Cu Peptide in the Light of the New Gene Data." International Journal of Molecular Sciences 2018;19(7):1987. DOI 10.3390/ijms19071987. Read at source. The authors are affiliated with R&D Skin Biology. This is cited here as the origin of the GHK-Cu hair claim and not as evidence for it, because the paper describes no human hair study and attaches no supporting study to the claim in its text. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17
5. Hasanzadeh H, Nasrollahi SA, Halavati N, Saberi M, Firooz A. "Efficacy and safety of 5% minoxidil topical foam in male pattern hair loss treatment and patient satisfaction." Acta Dermatovenerologica Alpina, Pannonica et Adriatica 2016;25:41-44. DOI 10.15570/actaapa.2016.12. Read at source. A before-and-after trial with no control group, enrolling 17 men with 14 completing 16 weeks and 12 completing 24 weeks, so it is the weakest of the human evidence cited here. ↩1 ↩2 ↩3
6. Singh S, Patil A, Kianfar N, Waśkiel-Burnat A, Rudnicka L, Sinclair R, Goldust M. "Does topical minoxidil at concentrations higher than 5% provide additional clinical benefit?" Clinical and Experimental Dermatology 2022;47(11):1951-1955. DOI 10.1111/ced.15338. PMID 35881665. Read at source. The article reports most trials without numeric values or p values, and its account of one study is internally inconsistent, with the text naming 5% where the table lists 15%. ↩1 ↩2 ↩3