If you own a copper peptide serum, you bought it on the strength of the research, and the research is real. GHK has been studied since 1973 and appears in more peer-reviewed literature than almost anything else in a bathroom cabinet. The question nobody puts on the box is the one that decides whether any of it matters: does the peptide in that bottle actually get anywhere?
GHK is at once one of the best-documented ingredients in cosmetic science and one of the worst-delivered. Those two facts are not in tension — together they are the whole story. A molecule can have five decades of laboratory work behind it and still spend most of its life on the surface of your face, because skin is very good at its job, and its job is keeping molecules like this one out.
GHK is a tripeptide — three amino acids in sequence, glycine-histidine-lysine. Loren Pickart isolated it from human blood plasma in 1973. It is not a synthetic invention but a fragment your body already makes and carries in plasma, saliva and urine.
Its defining property is that it binds copper, tightly and one-to-one, forming the complex written as GHK-Cu. Copper is not decoration here: it is a required cofactor for the enzyme that cross-links collagen and elastin. The peptide is, in effect, a very small copper delivery vehicle.
Circulating levels fall with age — roughly 200 ng/mL at age 20 and around 80 ng/mL by age 60, a figure reported consistently across Pickart's published reviews. That decline tracks the years when tissue repair slows, which is the observation that launched fifty years of interest. It is a correlation, and worth remembering as one.
The paper most often cited in copper peptide marketing is Pickart and Margolina's 2018 review in the International Journal of Molecular Sciences. Its headline finding, drawn from Broad Institute Connectivity Map data, is that GHK altered expression of about 31.2% of 13,424 human genes at a 50% change threshold — roughly 59% of them up, 41% down.
That number is striking and routinely misread. The Connectivity Map is a screening database built from cultured human cell lines exposed to a compound and profiled for gene-expression changes. It tells you a molecule is biologically busy. It does not tell you what happens in a living dermis, at what dose, or whether anything a person would notice follows.
The human topical evidence is thinner. The study almost every article leans on is Leyden, Stephens and Finkey's trial of a GHK-Cu cream in 71 women with photoaged skin, twice daily for 12 weeks, reporting improvements in skin density, laxity, clarity and fine-line depth. It was presented as a conference abstract at the 2002 American Academy of Dermatology meeting and, as far as we can establish, never published in full.
Thirteen thousand genes and a conference poster is an unusual shape for an evidence base.
The stratum corneum is the outermost fifteen to twenty layers of dead, flattened cells packed in a lipid matrix — bricks and mortar. It is a lipid environment, and that is precisely the problem, because GHK-Cu is charged and water-loving.
There is a rule of thumb in transdermal pharmacology, formalized by Bos and Meinardi in 2000: molecules above about 500 daltons do not passively cross intact skin in useful amounts. Copper peptide copy invokes it selectively, because the number cuts an awkward way. Free GHK weighs about 340 daltons; the GHK-Cu complex about 404. Both sit comfortably under the limit.
That threshold was always necessary, not sufficient. A small molecule still has to dissolve into a lipid barrier to cross it, and a charged hydrophilic complex does that poorly however light it is. Two problems stack on top: peptides are vulnerable to enzymes in the skin's upper layers, and copper complexes are chemically restless in a bottle and on a face alike.
Every serious attempt to use GHK on skin has been an attempt to solve delivery. There are four broad routes, and each gives something up.
Reformulate the vehicle. Liposomes, penetration enhancers, occlusion, or chemical modification of the peptide itself. Not hand-waving: a 2014 study found that adding a fatty-acid tail to the peptide KTTKS produced measurable distribution across all skin layers, where the unmodified peptide was undetectable. Formulation chemistry can change the answer — but that was a different peptide, and GHK-Cu has no equivalent, well-replicated fix in the public literature.
Open channels first, then apply. Microneedling and similar devices create controlled punctures, and actives go in afterward. Delivery improves. The trade is deliberate injury, downtime, operator dependence, and a regulatory point: FDA's aesthetic microneedling classification under 21 CFR 878.4430 expressly excludes devices intended for transdermal delivery of cosmetics, drugs or biologics.
Attach the peptide to something that makes its own way through. The route behind the spicule treatments now appearing in clinics — silica microstructures from a freshwater sponge, with GHK bonded to them by a tether designed to release once it meets glutathione inside skin. Every finding here carries its model, because the findings are early. On the benchtop, reduced glutathione accelerated release roughly fourfold versus buffer. In a 3D-bioprinted human skin model published in Micromachines in June 2026, the bonded version delivered more than five times as much peptide as a loose spicule-and-peptide mixture. In a 20-person split-face study published the same year, both sides got the spicule regimen — no untreated control — and both improved. No human collagen measurement exists. We walk through it in our explainer on what a Glassifier® actually is.
Bypass the barrier entirely. Inject it. Delivery stops being a question, and everything else — clinician oversight, a needle, and a human evidence base thinner than the internet implies — becomes one.
The honest answer is not "no." Something almost certainly gets in: Mazurowska and Mojski, in the Journal of Cosmetic Science in 2008, tracked copper-peptide complexes across a lipid membrane built to model the stratum corneum and found transport occurring. That is a model membrane, not a face, but it is not nothing. Delivery into the upper epidermis is plausible. Delivery to the fibroblasts in the dermis, at a dose resembling what the laboratory work used, is what nobody has shown for a consumer product.
Then the variables stack. Published human work used concentrations around 1–2%, and most serums do not disclose theirs. pH matters, and copper peptides want something near neutral. Packaging matters, because light and air degrade the complex. So does whether the formula carries the pre-formed complex or free peptide relying on ambient copper. So does barrier condition — compromised, post-procedure or simply dry skin lets more through, which is why the post-treatment setting is where delivery is least in doubt.
A well-formulated copper peptide serum is a reasonable thing to own and an unreasonable thing to expect much from on its own. Alongside sunscreen, a retinoid and whatever in-clinic work you are doing, it is defensible. As the load-bearing member of a routine, it is being asked to do something nobody has shown it can do.
You will be told not to layer copper peptides with direct vitamin C — L-ascorbic acid specifically. The advice is sound. The reasoning given for it is stated far more confidently than the evidence behind it.
The chemistry is well established in general terms: ascorbate reduces copper(II) to copper(I), and that redox pair generates reactive oxygen species while degrading the ascorbate. There is also a formulation mismatch, since L-ascorbic acid needs a pH near 3.5 to stay stable and absorb while copper peptide formulas sit closer to neutral.
What does not exist, as of 2026, is a published clinical study showing that layering them on a human face measurably reduces the benefit of either. The caution is an extrapolation from solution chemistry to a bathroom counter — a good rule with a thin evidence base, which is a different thing from a bad rule. Separating them costs nothing, so separate them: vitamin C in the morning, copper peptides at night. Stabilized derivatives such as tetrahexyldecyl ascorbate are much less reactive.
Naya offers GHK-Cu inside our peptide therapy program, prescribed by a clinician and given as a subcutaneous injection rather than applied to the surface. That choice is entirely about the problem described above. The molecule is the same molecule. The difference is whether it arrives.
The expectation gap runs in one direction. Patients arrive having used a serum for four or five months, describe the result as "maybe something," and conclude the ingredient does not work. The more accurate conclusion is narrower: that route did not deliver enough to judge the ingredient by. We are careful in the other direction too, because the human data on injected GHK for skin quality is genuinely limited, and patients deserve that before they start rather than after.
If you have a serum on your shelf and want a straight answer about whether it earns its place — or whether a delivery-assisted or prescribed route suits your skin better — that is exactly the kind of conversation we have at Naya Wellness, at our Chatsworth and Santa Barbara clinics or by California telehealth with Carolyn and our provider team.
Partially, and nobody can tell you how much for a specific product. Laboratory work using membranes modeling the stratum corneum shows copper-peptide transport occurs, but delivery into the dermis at a research-comparable dose has not been shown for any consumer serum.
Yes, at different times. Copper and L-ascorbic acid react with each other in solution and want opposite pH ranges, so the standard advice is vitamin C in the morning and copper peptides at night. That rests on chemistry rather than on any clinical study of the combination on skin, but separating them costs nothing.
The most-cited human topical study ran twice-daily application for 12 weeks, which is a fair benchmark. Anything promising visible change in days is describing hydration, not remodeling.
Some people report breakouts or irritation after starting them, often labeled "purging." That term was borrowed from retinoids and exfoliating acids and has not been characterized for copper peptides. Irritation, a reaction to something else in the formula, or coincidence are all at least as likely. Anything that spreads, burns or persists beyond a few weeks is worth reviewing with a clinician.
It depends on what you are treating and how much certainty you want about delivery. If your goal is general skin quality and the fundamentals are in place, a serum is a sensible low-stakes addition. If you want the delivery question answered rather than assumed, that is a candidacy conversation — and the honest version includes the limits of the evidence on every route.
Educational content, not medical advice. GHK-Cu within Naya's peptide therapy program is prescribed and compounded for individual patients; candidacy depends on your individual history and should be determined with a qualified clinician.