GHK-Cu serum vs injection vs retinol: differences, evidence and safety

Verdict: GHK-Cu serum is an unapproved cosmetic ingredient with small human skin studies; injected GHK-Cu is unapproved and has no published human trials; retinol is a cosmetic ingredient with a few small RCTs, and prescription tretinoin (Renova) is FDA-approved to help mitigate fine facial wrinkles. Retinoids hold by far the strongest evidence.

GHK-Cu is a copper-binding tripeptide found naturally in human plasma, and its level declines with age (Pickart 2015). A 2026 systematic review found 20 studies of GHK-Cu for aesthetic use, of which only 2 were randomized controlled trials [Human RCT] (Mokhtar 2026). Injected GHK-Cu has animal data only; FDA’s safety listing for compounded injectable GHK-Cu cites a risk of immune reactions and “limited data in humans” (FDA). Tretinoin cream (Renova) carries an FDA label for fine facial wrinkles [Approved label], with irritation in almost all trial patients (Renova label).

Item	GHK-Cu serum	GHK-Cu injection	Retinol / tretinoin
US FDA status	Not an approved drug; sold as a cosmetic, which needs no FDA approval (FDA)	Not approved for any use	Retinol: cosmetic, not approved. Tretinoin: approved prescription drug (Renova) for fine facial wrinkles (label)
Compounding (503A/503B)	Cosmetic serums are not compounded drugs. Compounded GHK-Cu for non-injectable routes is back in 503A Category 1 (under evaluation; interim, not approval) per FDA’s May 14, 2026 update (FDA)	Category 2 listing (injectable routes) withdrawn by nominators; injectable routes are excluded from the Category 1 entry (FDA); PCAC review due before end of February 2027, advisory only (FDA)	Tretinoin is available as an approved product
WADA 2026	GHK-Cu not named; WADA has not said whether S0 applies	Not named; S0 covers substances with no approval for human therapeutic use (WADA 2026)	Retinol and tretinoin not named
Strongest evidence	Human RCT, small (laser study n=13); larger studies not in PubMed-indexed journals	Animal only	Human RCTs (tretinoin n=296 and n=204; retinol n=36) and an approved label
Human trials	1 registered, not yet recruiting: NCT07706361	None registered	Five label trials for Renova (656 evaluable patients)
Last checked	2026-10-18	2026-10-07	2026-10-07

What changed (October 2026): a direct answer on GHK-Cu versus tretinoin is added to the questions below, and the athlete section now covers the 2027 WADA List.

GHK-Cu serum vs injection vs retinol at a glance

Feature	GHK-Cu serum	GHK-Cu injection	Retinol / tretinoin
What it is	Copper peptide in a topical cosmetic	Same peptide given under the skin	Vitamin A (retinol) or its active acid (tretinoin)
Class	Copper-binding tripeptide	Copper-binding tripeptide	Retinoid
Main proposed action	Collagen and matrix signaling (lab data)	Same, body-wide	Retinoic acid pathway; retinol acts as a precursor (Kang 1995)
Main research focus	Wrinkles, laser recovery, wounds	Wound healing in animals	Photoaging and natural skin aging
Skin penetration	Contested: low through intact skin in one model	Bypasses the skin barrier	Effects measured in human skin biopsies
Human evidence	2 RCTs in a 2026 review; cosmetic studies outside PubMed-indexed journals	None	Many RCTs; approved label
Approval status	Cosmetic ingredient	Unapproved	Tretinoin approved (Rx); retinol cosmetic
WADA	Not named	Not named; S0 may apply	Not named
Timelines in studies	12 weeks (cosmetic studies)	No human data	16 to 48 weeks; label trials 24 weeks
Main safety concerns	Few safety data; product quality varies	Immune reactions, impurities; “limited data in humans” (FDA)	Irritation, sun sensitivity; tretinoin not for use in pregnancy

What is GHK-Cu serum?

GHK-Cu (glycyl-L-histidyl-L-lysine bound to copper) is a three-amino-acid peptide that holds a copper ion. In fibroblast cultures it stimulated collagen synthesis [Lab] (Maquart 1988), which is why it appears in “copper peptide” serums and creams. Those products are sold as cosmetics; US law does not require FDA approval for cosmetics, but a product intended to remove wrinkles or raise collagen production counts legally as a drug (FDA). A 2025 review noted “a surprising absence of clinical studies” despite the ingredient’s wide use (Mortazavi 2025).

What is injected GHK-Cu?

Injected GHK-Cu is the same peptide given as a shot under the skin. It is not approved for any use, and no injectable GHK-Cu trial is registered on ClinicalTrials.gov. FDA placed GHK-Cu “for injectable routes of administration” on its Category 2 safety list, citing possible immune reactions from aggregation and peptide impurities and “limited data in humans”; the nominators later withdrew it, which does not make it Category 1 (FDA). Body-wide healing effects have been reported in rats, mice and pigs [Animal] (Pickart 2015).

What are retinol and tretinoin?

Retinol is vitamin A, sold over the counter in cosmetics. Tretinoin is all-trans retinoic acid, the active form, sold by prescription. On human skin, retinol produced thickening similar to retinoic acid without measurable retinoic acid or significant redness, which suggests skin converts retinol into the active acid [Human, n=10] (Kang 1995). Renova (tretinoin cream) is approved as an add-on for fine facial wrinkles in people using sun avoidance and skin care; its label states it “does not eliminate wrinkles, repair sun-damaged skin, reverse photoaging” [Approved label] (Renova label).

How they work differently

• GHK-Cu serum: applied to the surface; it must cross the skin barrier first. Results conflict: copper from GHK-Cu permeated human skin samples over 48 hours in one lab study [Lab] (Hostynek 2010), while almost no peptide crossed intact skin in 9 hours in another [Lab] (Li 2015).
• GHK-Cu injection: skips the barrier and enters the body directly. That is the reason given for injecting it, but no published human study has measured what an injection does to skin.
• Retinol: works as a precursor that skin turns into retinoic acid, with collagen-gene effects that tended to be smaller than tretinoin’s in a 4-week biopsy comparison (Kong 2016).
• Tretinoin: is the active acid itself, so no conversion step is needed; in a 24-week trial it thickened the outer skin layer [Human RCT, n=296] (Olsen 1992).

Evidence head-to-head

The table grades the best published evidence for each outcome. Retinoids lead on every outcome where both have been studied.

Outcome	GHK-Cu serum	GHK-Cu injection	Retinol / tretinoin
Fine wrinkles	Human RCT, small; review reports less wrinkle volume and depth than controls (Mokhtar 2026)	No data	Approved label (tretinoin); Human RCT for retinol, n=36 (Kafi 2007)
Overall photoaging	Cosmetic studies in 71 and 67 women, not in PubMed-indexed journals, reported in a Skin Biology-authored (review)	No data	Human RCT, n=30 (Weiss 1988) and n=99 (Griffiths 1995)
Collagen in biopsies	Pilot study: 70% of women vs 40% with retinoic acid, not in a PubMed-indexed journal, reported in a Skin Biology-authored (review)	No human data	Human RCT: procollagen marker up at 12 months (Kang 2005)
Laser recovery	Human RCT, n=13: no objective benefit (Miller 2006)	No data	Not assessed here
Chronic wounds	Human RCT: positive in diabetic foot ulcers (Mulder 1994); null in venous ulcers (Bishop 1992)	Animal only (review)	Not a studied use here
Long-term safety	No long-term trials	No human data	Human RCT, 2 years, n=204 (Kang 2005)

The best-known GHK-Cu skin figures, such as collagen gains in 70% of women versus 50% with vitamin C and 40% with retinoic acid after a month, and 12-week facial studies in 71 and 67 women, come from a review by authors at the company Skin Biology; the studies behind them are a 1998 pilot in a journal not indexed in PubMed, a 2002 meeting abstract and a 2005 book chapter, none of them trials in PubMed-indexed journals (Pickart 2015). A 2026 evidence map concluded that GHK-Cu remains “a promising but unproven therapeutic cargo” (Mateescu 2026).

Which fits which goal?

• Fine wrinkles from sun damage: tretinoin has the most evidence, including an FDA label. In a 16-week trial, 14 of 15 people treated on the face improved versus none on vehicle [Human RCT, n=30] (Weiss 1988).
• Naturally aged skin: retinol has one RCT; fine wrinkling scores fell by 1.64 points versus 0.08 with vehicle over 24 weeks in adults with a mean age of 87 [Human RCT, n=36] (Kafi 2007).
• Healing after a procedure: GHK-Cu serum raised patient satisfaction after laser resurfacing, but blinded evaluators found no difference in redness or wrinkles [Human RCT, n=13] (Miller 2006).
• Diabetic foot ulcers: a GHK-Cu gel improved median closure to 98.5% versus 60.8% with vehicle [Human RCT] (Mulder 1994). This was a medical wound gel, not a cosmetic serum.
• Injected GHK-Cu: no human evidence for any skin goal.

These are evidence statements, not recommendations. A dermatologist or other clinician is the person to decide which, if any, suits a given patient. Our method for grading evidence is explained on How We Compare Peptides.

Can they be combined?

No controlled human study has tested GHK-Cu together with retinol or tretinoin; the only direct comparison is a 1998 pilot, in a journal not indexed in PubMed, that tested them separately, not together (Pickart 2015). Tretinoin’s label warns that other topical products that irritate or dry the skin “should be used with caution” alongside it (Renova label). For other GHK-Cu blends, see KLOW vs GLOW vs GHK-Cu.

Safety and regulatory status

GHK-Cu serum. As a cosmetic it is not FDA-approved and needs no pre-market approval, though cosmetics must be safe in customary use (FDA). A microneedle study on skin cells and pig skin saw no obvious signs of irritation [Lab and animal] (Li 2015), but no long-term safety trial exists. Separately, FDA put compounded GHK-Cu “except for injectable routes of administration” back in 503A Category 1 in May 2026 after one nominator kept its non-injectable nomination; Category 1 means the substance is under evaluation and FDA does not intend to act against compounders who meet its conditions, not that it is approved (FDA Category 1 list).

GHK-Cu injection. FDA lists immunogenicity, aggregation and peptide-related impurities as risks and says human data are limited (FDA Category 2 page). That Category 2 listing was later withdrawn by the nominators, and the injectable route was left out when GHK-Cu returned to Category 1. Under section 503A, pharmacies may compound only from bulk substances that have a USP or NF monograph, are components of an FDA-approved drug, or appear on the 503A bulks list (FDA 503A page); GHK-Cu is still only being considered for that list. FDA’s Pharmacy Compounding Advisory Committee will discuss GHK-Cu for the 503A bulks list at a meeting “before the end of February 2027”; the date is not yet set and the committee’s votes are non-binding (FDA PCAC). See Are peptides safe? for the wider picture.

Retinol and tretinoin. In the Renova trials, 32% of patients had severe irritation, stopped temporarily or needed a mild steroid cream, and about 4% stopped because of side effects; the label warns of sunburn susceptibility and says not to use it in pregnancy, as oral tretinoin causes birth defects in animals [Approved label] (Renova label). Higher-strength tretinoin irritated more without working better [Human RCT, n=99] (Griffiths 1995).

Athletes. The WADA 2026 List does not name GHK-Cu, retinol or tretinoin. Its S0 class prohibits any pharmacological substance not covered elsewhere and with no current approval by any regulator for human therapeutic use (WADA 2026). WADA has not stated whether S0 applies to GHK-Cu, injected or topical; athletes should ask their anti-doping organization. The 2027 List, in force from 2027-01-01, names “peptides” in general in S0, which brings a peptide such as GHK-Cu closer to that class; retinol and tretinoin are not peptides and remain unaffected (USADA athlete advisory).

How long do effects take in studies?

Tretinoin trials reported facial improvement at 16 weeks (Weiss 1988) and ran to 48 weeks (Griffiths 1995); Renova’s label trials lasted 24 weeks, and some benefit may fade within 12 weeks of stopping (Renova label). Retinol reduced facial wrinkles after 12 weeks in an industry-run study (Kong 2016) and fine wrinkling at 24 weeks in older adults (Kafi 2007). The GHK-Cu cosmetic studies ran 12 weeks (Pickart 2015). There is no human timeline for injected GHK-Cu.