Glycolic Acid Studies: The Clinical Evidence in One Table
This page lists 122 human studies in which glycolic acid was applied to skin, published between 1993 and 2026 and indexed in PubMed. We read every one from its abstract on 22 November 2026 and recorded the same fields for each: who took part, where on the body, home or clinic, strength and pH, schedule, what it was compared with, what was found and what went wrong.
Of the 122, 60 are clinic peels, 41 used products at home, 11 combined a clinic peel with a home product, and 9 were test protocols in which investigators applied the acid. 50 tested glycolic acid only together with another active ingredient, so on their own they cannot show what glycolic acid does. 49 describe themselves as randomized. Results that do not flatter glycolic acid stay in the table.
What the studies show, in six points
1. The best home-strength evidence is small, and mostly on the face
Three placebo-controlled trials carry most of the weight for everyday products. A 5% unneutralized cream used daily for 3 months on the face and neck improved texture and discoloration more than placebo in 75 people; the wrinkle trend was not significant (Thibault 1998). An 8% cream used twice daily for 22 weeks gave at least one grade of improvement in facial photodamage to 76% of users against 40% on the vehicle cream (Stiller 1996). A 10% emulsion at pH 4, used every evening, improved mild acne more than placebo in 120 people by day 45 (Abels 2011, run by the product's maker). At pH 4 that 10% works out to about 4% free acid by the formula in our percentage guide.
2. Glycolic acid makes skin more sensitive to the sun, and one older study says otherwise
Three studies found more UV damage on glycolic-treated skin. Four weeks of a 10% product at pH 3.5 raised sunburn-cell counts and lowered the dose needed to burn, with the effect gone a week after stopping (Kaidbey 2003); 10% glycolic increased redness, DNA damage and sunburn cells after simulated sunlight while 2% salicylic acid did not (Kornhauser 2009, FDA authors); and glycolic-treated skin tanned more after UVB (Tsai 2000). A 1996 study reported the opposite, an SPF of about 2.4 after three weeks of 8% products (Perricone 1996). We show both and do not average them; our sun exposure guide covers what to do about it.
3. Head to head, glycolic acid mostly ties, and sometimes loses
Against Jessner's solution, salicylic, salicylic–mandelic, tretinoin and retinol peels and trichloroacetic acid, glycolic peels usually performed about the same. Tretinoin or retinoic acid beat it in three comparisons: on facial elasticity (Piérard 1996), on aged forearm skin, where 10% glycolic showed no significant effect (Piérard 1999), and on underarm acanthosis nigricans (Ghiasi 2024). Amino fruit acid, salicylic–mandelic and lactic peels were better tolerated in their trials. See our glycolic acid vs retinol and glycolic acid vs salicylic acid comparisons.
4. As an add-on for melasma, the record is mixed
Adding a series of glycolic peels to a home cream gave faster or greater melasma improvement in several trials (Sarkar 2002, Chaudhary 2013, Sarkar 2017), but four peels added nothing to 4% hydroquinone in a split-face trial (Hurley 2002). At home strength, adding 10% glycolic acid to 4% hydroquinone improved melasma less markedly than hydroquinone alone and caused the most side effects of five groups (Ibrahim 2015). Our dark spots guide covers the options.
5. Outside the face, the evidence is thin
75 of the 122 studies treated the face; 24 do not say where. Home-strength trials elsewhere on the body are few: forearm skin in photoaging studies, a 20% cream combined with tretinoin or vitamin C on white stretch marks of the abdomen and thighs (Ash 1998), and a 10% scalp lotion, used twice daily for 8 weeks, that improved scalp psoriasis in every one of 20 patients (Kostarelos 2000). Keratosis pilaris on the arms and thighs improved with 50–70% clinic peels, but the 9 people followed up at 5 years were back where they started (Tian 2021). No study we found sets a home-use strength for any body area, which is why our body and underarm guide labels its ranges as working guidance.
6. Most studies never report pH
Only 11 of the 122 abstracts give a pH, so the share of the acid that is actually free to act can be estimated for just 8 of them (shown in the table). That matters: peel solutions below pH 2 caused crusting and necrosis that partially neutralized solutions above pH 2 did not (Becker 1996). Most product labels have the same gap; our statistics page counts how few brands publish pH.
The 122 studies
Newest first. Filter by topic, body area or setting, or search any word. Each study links to its PubMed record, and the terms used, from split-face to MASI score, are defined in our glycolic acid glossary.
| Study | Design | Where and setting | Glycolic acid | Schedule | Compared with | What was found | Side effects |
|---|---|---|---|---|---|---|---|
| Kwon 2026 Industry author | Randomized, double-blind, split-face, n = 37 | Face · Home use | 10% (with 0.1% lecithin), pH 3.9 (≈ 4.6% free acid) · biphasic serum | Not stated · 2 weeks | conventional formulation | The biphasic serum improved exfoliation measures by 45.52% and improved brightness, texture and hydration. | Redness and water loss improved. |
| Rini 2026 | Randomized controlled, n = 60 | Face · Clinic peel | 35% · peel | six sessions, 2 weeks apart | 4% retinol peel | Melasma improved in both groups with no difference. | Mild side effects in 3.3% (glycolic) vs 13.3% (retinol). |
| Sun 2026 Combination | Prospective split-face, n = 20 | Face · Clinic peel | Not stated · peel plus intense pulsed light | six monthly sessions | intense pulsed light alone | The combined side improved spots and redness after one session, where light alone needed three. | Not reported in the abstract |
| Pan 2025 Combination | Randomized, two groups, n = 40 | Face · Clinic peel | Not stated · peel plus light and oral medicine | glycolic peels alternated weekly with intense pulsed light · 4 months | oral tranexamic acid and intense pulsed light without peels | Melasma scores improved 60.70% with glycolic added vs 46.81% without. | Not reported in the abstract |
| Alajmi 2024 Combination | Evaluator-blinded, split-hand, n = 20 | Hands · Clinic peel | 3% (with 15% trichloroacetic acid) · peel | three treatments, 4 weeks apart · 12 weeks after last | untreated hand | Hand age spots improved significantly on the treated hand. | No adverse events; mean pain 3.8 of 10. |
| Ghiasi 2024 | Randomized, single-blinded, split-side, n = 30 | Underarms, neck · Clinic peel | 70% · peel | every 2 weeks, four sessions · 8 weeks | 0.05% tretinoin cream every other night | For acanthosis nigricans, tretinoin worked better on the underarms; neither treatment worked significantly on the neck. | Not reported in the abstract |
| Moghimipour 2024 | Comparison of two gels | Not stated · Home use | not stated · liposomal gel vs plain gel | Not stated | plain glycolic gel | The liposomal glycolic gel reduced hyperpigmentation, fine wrinkling and lentigines more than the plain gel. | Fewer side effects with the liposomal gel. |
| Chen 2022 | Open-label, uncontrolled, n = 30 | Face · Home use | 5% · gel essence | Not stated · 4 weeks | none | Acne lesions and severity scores improved; redness and melanin readings fell by week 4. | Not reported in the abstract |
| Dayal 2022 Combination | Randomized controlled, n = 40 | Face · Clinic peel | 35% · peel plus microneedling | every 2 weeks | microneedling plus 15% trichloroacetic acid | Acne scars improved equally in both groups; texture improved more with glycolic. | Not reported in the abstract |
| Liu 2022 | Randomized, evaluator-blinded, split-face, n = 30 | Face · Clinic peel | 35% · peel | three peels, 2 weeks apart | adapalene gel nightly | Flat facial warts fell by more than 70% in 15 patients (50%) on the glycolic side vs 8 on the adapalene side. | Mild redness and peeling. |
| Abdel-Majid 2021 Combination | Split-face comparison, n = 30 | Face · Clinic peel | 70% (with 20% trichloroacetic acid) · peel | Not stated | modified Jessner's plus 20% trichloroacetic acid | Melasma improved on both sides, more on the Jessner's side. | Not reported in the abstract |
| Campione 2021 Combination | Assessor-blinded, uncontrolled, n = 12 | Face · Home use | 4% (with 0.02% retinoic acid) · gel | three times a week · 4 weeks | none | In severe skin aging, dark spots fell 40% and wrinkle severity 12% by week 4. | Not reported in the abstract |
| Campos 2021 Combination Industry author | Open-label, n = 42 | Face · Home use | not stated (with dioic, salicylic, capryloyl salicylic and citric acids) · serum | daily · 56 days | none | Post-acne dark marks fell 29.4% in number and oiliness fell 30.7% by day 7. | Not reported in the abstract |
| DiNatale 2021 Combination Industry author | Two randomized, double-blind, split-face trials plus lab work | Face · Home use | not stated (phytol–glycolic product rotated with retinol) · rotational regimen | Not stated · 6 and 12 months | untreated side; marketed products | Rotating retinol with a phytol–glycolic product improved crow's-feet wrinkles more than either comparison. | Not reported in the abstract |
| Houshmand 2021 Combination | Open-label | Face · Home use | 10% (with 2% phytic acid and a soothing complex) · emulsion | Not stated · 12 weeks | none | Post-inflammatory hyperpigmentation, discoloration, texture and tone improved from week 4. | Well tolerated. |
| Lukic 2021 | In vivo skin pH study | Not stated · Test protocol | 2%; 10% · cream | single and repeated application · up to 7 days | vehicle | Skin pH read 3.89–4.25 five hours after application and stayed at 5 or below over 7 days, more so with 10%. | No barrier impairment or irritation reported. |
| Sahu 2021 | Comparative, three groups, n = 90 | Face · Clinic peel | 30% · peel | every 2 weeks · 12 weeks | 92% lactic acid peel; 15% trichloroacetic acid peel | Glycolic and trichloroacetic peels improved melasma equally and more than lactic; quality of life improved most with glycolic. | Adverse effects most with trichloroacetic acid, then glycolic. |
| Sitohang 2021 | Randomized, double-blind, split-face, n = 126 | Face · Clinic peel | 35% · peel | single peel | sodium bicarbonate neutralizer vs water | Pain right after neutralizing differed between sides, but redness and itch did not; the authors conclude water can replace bicarbonate in skin types IV–V. | Not reported in the abstract |
| Tian 2021 | Uncontrolled, with 5-year follow-up, n = 25 | Arms or thighs (keratosis pilaris sites) · Clinic peel | 50% or 70% · peel | four sessions, on days 0, 20, 40 and 60 · 80 days; 5-year follow-up | none | Keratosis pilaris bumps improved by 8%, 12%, 36% and 60% at days 20, 40, 60 and 80; at 5 years the 9 people followed up were no different from before treatment. | Not reported in the abstract |
| Heuser 2020 Combination | Randomized, within-patient, n = 17 | Arms · Clinic peel | 70% (followed by 5% 5-fluorouracil) · peel then cream | peel every 15 days | 5% 5-fluorouracil cream alone | Actinic keratoses on the arms fell 75% with peel plus 5-fluorouracil and 85.71% with the cream alone, not significantly different. | Not reported in the abstract |
| Kubiak 2020 Combination | Comparison of two peels, n = 40 | Face · Clinic peel | 70% (with 15% trichloroacetic acid) · peel | five sessions, 14 days apart | 35% trichloroacetic acid alone | Both improved elasticity, hydration, melanin and redness; the glycolic combination hydrated more and lowered melanin more, while 35% trichloroacetic acid reduced wrinkles more. | No adverse events needing intervention. |
| Villani 2020 Combination | Prospective, uncontrolled, n = 25 | Face · Home use | not stated (with retinoids, a peptide, salicylic acid and niacinamide) · gel | once daily in the evening · 8 weeks | none | Acne lesions fell 80% by week 8; water loss did not change. | No redness or dryness reported. |
| Zdrada 2020 Combination | Split-face comparison, n = 14 | Face · Clinic peel | not stated (with salicylic acid) · peel | four treatments, 2 weeks apart | 50% pyruvic acid | Both sides changed in hydration and colour; the authors judge the two peels comparable. | Not reported in the abstract |
| Garg 2019 | Randomized, three groups, n = 30 | Face · Clinic peel | 35% · peel | four peels, every 15 days | 35% glycolic plus 10% or 20% trichloroacetic acid spot peel | Melasma improved in all groups with no difference; adding the spot peel gave no extra benefit. | Fewest side effects with glycolic alone. |
| Saadawi 2019 | Randomized, three groups, n = 30 | Face · Clinic peel | Not stated · peel | six sessions, 2 weeks apart | microneedling; both combined | Acne scars improved in all groups, most with peel plus microneedling. | Not reported in the abstract |
| In Jae 2018 Combination | Randomized, evaluator-blind, split-face, n = 20 | Face · Clinic peel | 50% buffered (with 0.5% salicylic acid), pH 3.0 (≈ 43.6% free acid) · peel | two sessions, 2 weeks apart | Jessner's solution | Acne lesions fell equally on both sides. | Fewer side effects on the glycolic side. |
| Khalil 2018 Combination | Case series, n = 15 | Not stated · Home use | 10–20% (with lovastatin 2% and cholesterol 2% cream) · cream | Not stated · 3 months | none | In congenital ichthyosis, severity scores fell 57.5% at 3 months. | Mild irritation and burning. |
| Rouvrais 2018 Industry author | Randomized controlled, n = 55 | Face · Clinic peel | 20%, 50%, 70% in sequence · peel | three peels, 2–3 weeks apart · 8 weeks | 0.1% retinaldehyde cream daily | Both reduced crow's-feet wrinkles similarly; the cream refined texture better. | Peels caused about 12 times more physical signs of irritation. |
| Zaki 2018 | Split-neck comparison, n = 20 | Neck · Clinic peel | 70% · peel | three sessions | fractional CO2 laser | Pseudo-acanthosis nigricans on the neck improved 43% on the glycolic side vs 19% on the laser side. | Not reported in the abstract |
| Bulbul Baskan 2017 | Comparative, three groups, n = 30 | Not stated · Clinic peel | 50% · peel | timed to menstrual cycle phase · 3 months | the same peel at other cycle phases | Acne improved in all three groups but significantly only when peels were done on cycle days 10–14. | Not reported in the abstract |
| Dayal 2017 Combination | Comparative, two groups, n = 60 | Face · Home and clinic | Not stated · peels plus home cream | peel every 3 weeks with 20% azelaic acid cream twice daily · 24 weeks | azelaic acid cream alone | Adding peels improved melasma scores from week 12 on and improved quality of life. | Minor reversible side effects in both groups. |
| Rana 2017 Combination | Randomized controlled, n = 60 | Face · Clinic peel | 70% · peel plus microneedling | peels at weeks 3, 9 and 15 between microneedling sessions · 22 weeks | microneedling alone | Acne-scar scores fell more with peels added (39.65 vs 29.58), and texture improved more. | Not reported in the abstract |
| Sarkar 2017 Combination | Randomized, two groups, n = 30 | Face · Home and clinic | Not stated · peels plus home cream | serial peels with modified Kligman cream · 21 weeks | modified Kligman cream alone | Post-inflammatory hyperpigmentation scores were significantly better with peels added at 12 and 21 weeks, in skin types III–V. | Side effects in both groups, managed with emollients; no dropouts. |
| Dayal 2016 | Comparative, three groups, n = 90 | Under-eye area · Clinic peel | 20% · peel | every 3 weeks · 12 weeks | 15% lactic acid peel; 20% vitamin C daily | More than half the dark circles faded in 73.34% (glycolic), 56.67% (lactic) and 26.67% (vitamin C). | Adverse effects were most frequent with glycolic. |
| Ichiyama 2016 | Case report, n = 2 | Not stated · Clinic peel | Not stated · peel | every 2 weeks | none | In inherited generalized acanthosis nigricans, darkening and coarseness improved with regular glycolic peels. | Few adverse effects. |
| Sarkar 2016 | Randomized, three groups, n = 90 | Face · Clinic peel | 35% · peel | every 14 days after 4 weeks' priming with hydroquinone and tretinoin · 12 weeks | salicylic–mandelic peel; phytic acid combination peel | Melasma scores fell 62.36% (glycolic), 60.98% (salicylic–mandelic) and 44.71% (phytic); glycolic matched salicylic–mandelic and beat phytic. | The salicylic–mandelic peel was better tolerated. |
| Wakade 2016 | Split-face comparison, n = 40 | Face · Clinic peel | increasing concentrations · peel | four treatments, 3 weeks apart | monopolar radiofrequency | Biopsy and ultrasound showed dermal improvement in 35 of 40 patients on both sides; the two treatments worked equally. | Not reported in the abstract |
| Behm 2015 | Uncontrolled test in 10 diabetic subjects; 28-day half-side trial in 30 | Not stated · Home use | not stated, pH 4 · water-in-oil emulsion | not stated · 28 days | untreated half | The pH 4 glycolic emulsion lowered skin surface pH in healthy, elderly and diabetic subjects. | Not reported in the abstract |
| Ibrahim 2015 Combination | Five-group comparison, n = 100 | Face · Home use | 10% (with 4% hydroquinone, with or without 0.01% hyaluronic acid) · cream | Not stated | 4% hydroquinone alone; with hyaluronic acid; placebo | The hydroquinone-plus-glycolic group improved less markedly than hydroquinone alone and reported the most side effects; adding hyaluronic acid as well worked well. | Side effects most frequent in the hydroquinone-plus-glycolic group. |
| Mekas 2015 | Double-blind, prospective, three arms, n = 75 | Face · Home use | 4%; 8% (with 2% citric acid) · cream | twice daily · 12 weeks | hydrolyzed roe protein cream | The 4% cream improved fine lines; the 8% glycolic-citric cream improved clarity, roughness and hyperpigmentation. | More stinging and burning than the roe cream. |
| Vachiramon 2015 Combination | Randomized split-face, n = 15 | Face · Clinic peel | 30% · peel plus laser | five weekly sessions · 12 weeks | low-fluence Q-switched Nd:YAG laser alone | In men with melasma, the combined side lightened up to 52.3% at week 4 but relapsed by weeks 8–12; the authors judged the side effects not worth it. | One patient (8.3%) had guttate hypopigmentation that had not resolved at 12 weeks. |
| Bhatia 2014 Combination | Open-label, n = 25 | Not stated · Home use | 10% buffered (with 1% salicylic acid) · cleanser plus treatment gel | twice daily · 6 weeks | none | Inflammatory lesions fell 98.55% and non-inflammatory lesions 56.10% by week 6. | Very well tolerated. |
| Kaminaka 2014 | Randomized, double-blind, placebo-controlled, split-face, n = 26 | Face · Clinic peel | 40%, pH 2.0 (≈ 39.4% free acid) · peel | five sessions, 2 weeks apart | placebo | Moderate acne lesions fell significantly more on the glycolic side, especially non-inflamed lesions; sebum fell on both sides alike. | Not reported in the abstract |
| Kubiak 2014 | Comparison of two peels, n = 20 | Face · Clinic peel | 70% · peel | five sessions, 14 days apart | 15% trichloroacetic acid peel | Both peels improved elasticity and hydration; melanin fell significantly only after glycolic, redness rose only after trichloroacetic acid. | Not reported in the abstract |
| Masini 2014 Combination | Retrospective, n = 30 | Not stated · Home use | 6% (with 0.1% retinaldehyde and 0.1% efectiose) · cream | once daily in the evening, with SPF 50 by day · 8 weeks | none | Over a sunny season, the acne score fell from 1.83 to 0.90 at 60 days. | Side effects very uncommon. |
| Baumann 2013 Combination | Split-face comparison | Face · Home use | not stated (with salicylic, capryloyl salicylic, citric and dioic acids) · multi-acid product | Not stated | 1% clindamycin plus 5% benzoyl peroxide gel | Tolerance and acne results were similar for the two products. | Not reported in the abstract |
| Chaudhary 2013 Combination | Comparative, two groups, n = 40 | Face · Clinic peel | Not stated · peel | serial peels with hydroquinone, hydrocortisone and tretinoin cream · 24 weeks | the cream alone | Adding peels gave earlier and greater melasma improvement. | Not reported in the abstract |
| Kakudo 2013 | Uncontrolled, n = 23 | Face · Clinic peel | Not stated · peel | five peels, 2 weeks apart | none | Areas of light pigmentation shrank significantly; darker pigment did not change significantly. | Not reported in the abstract |
| Prestes 2013 | Randomized, three groups, n = 27 | Outer eye area · Clinic peel | 70% · peel | monthly, three applications · 3 months | 85% lactic acid peel; sunscreen only | Both peels reduced fine wrinkles at the outer eye after three applications; the sunscreen-only group did not change. | Not reported in the abstract |
| Ramírez 2013 Combination | Open-label, multicenter, n = 33 | Face · Home use | 20% buffered lotion (with 0.01% decapeptide-12 cream and SPF 30) · skin-care system | Not stated · 16 weeks | none | Melasma scores fell 60% by week 16 in Hispanic women. | No adverse events reported. |
| Kar 2012 | Randomized, three groups, n = 75 | Face · Clinic peel | Not stated · peel | every 2 weeks · 12 weeks | low- and high-fluence Q-switched Nd:YAG laser | All groups improved; low-fluence laser beat the peel, and the peel beat high-fluence laser. | Not reported in the abstract |
| Sobhi 2012 | Single-blind split-face, n = 14 | Face · Clinic peel | 70% · peel | six sessions | vitamin C by iontophoresis | Melasma improved on both sides, more on the vitamin C side, in skin types IV–V. | Few, transient side effects. |
| Takenaka 2012 | Lab and clinical | Cheeks · Clinic peel | 35%, pH 1.2 (≈ 34.9% free acid) · peel | single peel | untreated skin | Peeled cheeks carried fewer acne bacteria (propionibacteria) than untreated controls. | Not reported in the abstract |
| Abels 2011 Industry author | Randomized, double-blind, placebo-controlled, n = 120 | Not stated · Home use | 10%, pH 4 (≈ 4% free acid) · oil-in-water emulsion | once daily, evening · 90 days | placebo emulsion | Mild acne improved significantly versus placebo, with a difference by day 45. | No difference in adverse effects versus placebo. |
| Faghihi 2011 | Randomized, double-blind, split-face, n = 63 | Face · Clinic peel | 70% · peel | four sessions, 2 weeks apart | 1% tretinoin peel | Both sides improved similarly; discomfort was significantly higher on the glycolic side. | Rare complications on both sides. |
| Park 2011 Combination | Randomized, observer-blinded, split-face, n = 16 | Face · Clinic peel | 30% · peel plus laser | three peels, 2 weeks apart, with six weekly laser sessions | laser alone | Mixed melasma improved 37.4% on the combined side vs 16.7% with laser alone (pigment readings 32.6% vs 22%). | Not reported in the abstract |
| Perić 2011 | Uncontrolled safety study, n = 90 | Not stated · Clinic peel | different concentrations · peel | Not stated · 8 weeks | none | Glycolic peels for acne were well tolerated; at the end, 6 of 90 had strong redness, 10 strong peeling and 11 strong tightness. | Not reported in the abstract |
| Sharad 2011 Combination | Comparative, two groups | Face · Clinic peel | 35% · peel plus microneedling | microneedling every 6 weeks, five sessions | microneedling alone | Adding the peel improved scars, texture and post-acne pigmentation more than microneedling alone in skin types III–IV. | No added morbidity. |
| Draelos 2010 Combination | Randomized, double-blind, n = 80 | Face · Home use | not stated (with kojic acid and emblica extract) · formulation | twice daily · 12 weeks | 4% hydroquinone cream | Lightening matched 4% hydroquinone in mild to moderate facial discoloration. | Not reported in the abstract |
| Ilknur 2010 | Randomized, single-blind, split-face, n = 24 | Face · Clinic peel | rising concentrations · peel | 12 peels every 2 weeks · 6 months | amino fruit acid peel | Both peels improved comedonal acne with no difference between them. | The amino fruit acid peel was less irritating. |
| Ilknur 2010 | Randomized, single-blind, split-face | Face · Clinic peel | Not stated · peel | 12 peels every 2 weeks · 6 months | amino fruit acid peel | Melasma scores fell on both sides with no difference between them. | The amino fruit acid peel was less irritating. |
| Kumari 2010 | Comparative, two groups, n = 40 | Face · Home and clinic | 12% cream; 20–35% peels · home cream then peels | cream nightly for 2 weeks, then peels every 15 days · 12 weeks | 10–20% trichloroacetic acid peels | Melasma scores fell 79% with glycolic and 73% with trichloroacetic acid, not significantly different. | Fewer side effects with glycolic. |
| Rizza 2010 | In vivo comparison | Not stated · Test protocol | 10% · not stated | Not stated | mandelic acid; grape juice acid mixture | Glycolic acid exfoliated faster but caused more redness and more sun sensitivity than mandelic acid or the grape acid mixture. | More intense redness. |
| Tasic-Kostov 2010 | Comparison of test products, n = 77 | Not stated · Test protocol | not stated · gel; emulsion | Not stated · 2–4 weeks | 6% lactobionic acid | Lactobionic acid products irritated less and did not impair the skin barrier, compared with glycolic products. | Not reported in the abstract |
| Dreno 2009 Combination | Open, multicentre, uncontrolled, n = 397 | Not stated · Home use | 6% (with 0.1% retinaldehyde) · cream | alone or with usual acne treatments · 90 days | none | In women aged 30–40, inflamed and non-inflamed acne lesions fell by day 90. | Complaints rare. |
| Garg 2009 | Comparative, two groups, n = 44 | Face · Clinic peel | 35% · peel | every 2 weeks, six sessions | 20% salicylic–10% mandelic acid peel | Both worked; the salicylic–mandelic peel did better on active acne and post-acne pigmentation, with fewer side effects. | Not reported in the abstract |
| Kornhauser 2009 | Within-subject, vehicle-controlled, n = 14 | Back · Test protocol | 10% · cosmetic formulation | daily, Monday to Friday · 3.5 weeks | 2% salicylic acid; vehicle; untreated | Glycolic acid increased redness, DNA damage and sunburn cells after simulated sunlight; salicylic acid did not. | Not reported in the abstract |
| Bertin 2008 Combination Industry author | Randomized, double-blind, placebo-controlled, split-face, n = 40 | Face · Home use | not stated (with retinol and lactose) · cream | twice daily · 12 weeks | placebo cream | The treated side of the face improved significantly versus baseline and versus the placebo side. | Not reported in the abstract |
| Kessler 2008 | Randomized, double-blind, split-face, n = 20 | Face · Clinic peel | 30% · peel | every 2 weeks, six treatments | 30% salicylic acid peel | Both peels reduced acne equally by the second session; only the salicylic side stayed improved 2 months after. | More adverse events with glycolic after the first treatment. |
| Oresajo 2008 Industry author | Randomized split-face, n = 50 | Face · Clinic peel | 20–50% · peel | every 2 weeks · 12 weeks | 5–10% capryloyl salicylic acid peel | Fine lines improved in 30% (glycolic) vs 41% (salicylic derivative) and pigmentation in 34% vs 46%; the difference was not significant. | Not reported in the abstract |
| Rendon 2008 Combination | Pilot, n = 20 | Face · Home and clinic | Not stated · peels plus triple-combination cream | five peels alternating with cream · 12 weeks | none | Melasma was clear or almost clear in 13 of 20 (65%) by week 12. | Well tolerated. |
| Suswardana 2008 | Two-step randomized, double-blind trial, 20 per step | Not stated · Test protocol | 50% · solution | 4 minutes before a numbing cream | placebo pretreatment | Pretreating skin with glycolic acid brought adequate numbing at 30 minutes instead of 45. | Not reported in the abstract |
| Dreno 2007 Combination | Randomized, double-blind, vehicle-controlled, n = 145 | Not stated · Home use | 6% (with 0.1% retinaldehyde) · cream | every evening · 3 months | vehicle cream | In patients who used it as directed, acne scarring improved more than with vehicle, through less redness and pigmentation. | Not reported in the abstract |
| Erbil 2007 Combination | Randomized controlled, n = 28 | Face · Home and clinic | Not stated · peels plus home creams | serial peels with 20% azelaic acid and 0.1% adapalene · 20 weeks | azelaic acid and adapalene alone | Recalcitrant melasma improved more with peels added (P=0.048). | Three peel patients had mild post-inflammatory hyperpigmentation that cleared. |
| Sezer 2007 Combination | Randomized, split-hand, n = 25 | Hands · Clinic peel | 70% (with 35% trichloroacetic acid) · focal peel | single treatment · 2 months | cryosurgery | Hand lentigines cleared in 4 of 23 (17.4%) with the peel vs 5 of 23 (21.7%) with cryosurgery, not significantly different. | Not reported in the abstract |
| Salles 2006 Combination | Controlled, n = 77 | Around the mouth · Home use | 5–7% (with 0.01–0.05% tretinoin) · cream | Not stated · 3 months | untreated controls | In burn scars around the mouth, mouth-opening distances increased significantly versus controls. | Not reported in the abstract |
| Yamamoto 2006 | Histology comparison | Not stated · Test protocol | Not stated | once daily · 6 weeks | lactic, citric and acetic acids | Glycolic- and lactic-treated skin had less melanin in the basal layer than citric- or acetic-treated skin; collagen I rose with glycolic, lactic and citric acids. | Not reported in the abstract |
| Dréno 2005 Combination | Open, uncontrolled, n = 1709 | Not stated · Home use | 6% (with 0.1% retinaldehyde) · cream | with patients' usual acne treatments · 90 days | none | Well tolerated alongside other acne treatments; inflamed and non-inflamed lesions fell from day 0 to day 90. | Complaints rare. |
| Horikoshi 2005 Industry author | Mechanism study | Not stated · Clinic peel | 50%, pH 0.9 (≈ 49.9% free acid) · solution | single 30-minute application · 3 weeks follow-up | none | Skin surface pH fell to about 2 at 30 minutes and stayed near 3 for a day; the authors link shedding to an enzyme (cathepsin D) activated around pH 3. | Not reported in the abstract |
| Poli 2005 Combination | Randomized, double-blind, vehicle-controlled, multicentre | Not stated · Home use | 6% (with 0.1% retinaldehyde) · cream | Not stated · 3 months | vehicle cream | Marked global improvement at month 2 in 50.0% vs 26.3% on vehicle. | One patient stopped treatment. |
| Edison 2004 Industry author | Comparison of two regimens | Face · Home use | not stated · regimen of products | Not stated · 12 weeks | polyhydroxy acid (gluconolactone) regimen | Both regimens improved photoaging; the glycolic regimen did better on sallowness and skin recoil at week 12. | Stinging and burning were significantly worse with the glycolic regimen. |
| Khunger 2004 | Open split-face pilot, n = 10 | Face · Clinic peel | 70% · peel | weekly · 12 weeks | 1% tretinoin peel | Melasma improved on both sides with no difference between them. | Not reported in the abstract |
| Ziolkowski 2004 Combination | Clinical series with a control group, plus lab work, n = 12 | Not stated · Clinic peel | 5% (with 5-aminolevulinic acid) · ointment | before light treatment, one to three sessions | a control group (details not in the abstract) | Adding glycolic acid to the light-sensitizing ointment raised porphyrin levels in lab work; all 12 skin cancers (squamous cell carcinoma) regressed. | Not reported in the abstract |
| Fuchs 2003 | Randomized, double-blind, split-face, vehicle-controlled, n = 65 | Face · Home use | 15% · cream | Not stated · 6 months | 0.01% estradiol cream; the two combined; vehicle | Biopsies showed epidermal thickness up 27% with glycolic acid, 23% with estradiol and 38% with both. | Not reported in the abstract |
| Guevara 2003 Combination | Randomized controlled, n = 39 | Face · Home use | 10% buffered (with 4% hydroquinone, vitamins C and E, sunscreen) · cream | twice daily · 12 weeks | sunscreen cream | Melasma improved in 15 of 20 (75%) on the combination cream vs 2 of 15 (13%) on sunscreen alone. | Irritation was more common, and settled with short pauses and moisturizer. |
| Kaidbey 2003 | Randomized, double-blind, placebo-controlled, n = 29 | Back · Test protocol | 10%, pH 3.5 (≈ 6.8% free acid) · not stated | once daily, 6 days a week · 4 weeks | placebo | After UV exposure, glycolic-treated skin formed more sunburn cells and burned at a lower dose; the effect was gone a week after stopping. | Not reported in the abstract |
| Alam 2002 | Randomized split-face, unblinded, n = 10 | Face · Clinic peel | 20% · peel | weekly · 6 weeks | microdermabrasion | 7 of 10 patients preferred the peel side, but investigators saw no significant improvement from baseline with either. | No serious side effects. |
| Hurley 2002 | Prospective split-face, n = 21 | Face · Clinic peel | 20–30% · peel | every 2 weeks to one side, with 4% hydroquinone to both · 8 weeks | 4% hydroquinone alone | Adding four glycolic peels gave no extra lightening over 4% hydroquinone alone. | Not reported in the abstract |
| Sarkar 2002 Combination | Comparative, two groups, n = 40 | Face · Clinic peel | Not stated · peel | serial peels with modified Kligman cream · 21 weeks | modified Kligman cream alone | Melasma scores fell in both groups; adding peels gave faster and greater improvement in dark skin. | Few side effects in the peel group. |
| Bernstein 2001 Industry author | Vehicle-controlled | Forearms · Home use | 20% · lotion | twice daily · 3 months | vehicle lotion | Biopsies showed more hyaluronic acid and more type I collagen gene expression in glycolic-treated skin than vehicle-treated skin. | Not reported in the abstract |
| Funasaka 2001 | Uncontrolled imaging study | Face · Clinic peel | 35%, 50% or 70% · peel | Not stated | none | Total wrinkle length fell, most in people in their 50s and least over 70; the three strengths did not differ. | Not reported in the abstract |
| Javaheri 2001 | Uncontrolled, n = 25 | Face · Home and clinic | 10% lotion; 50% peel · home lotion then peel | lotion nightly for 2 weeks; peel monthly ×3 · 6 months | none | Melasma severity scores fell in 91% of women; epidermal melasma responded better than mixed. | No significant side effects. |
| Kharfi 2001 | Randomized, double-blind, n = 40 | Face · Home use | 12% · cream | Not stated · 45 days | 0.05% retinoic acid cream | 90% of glycolic users and 83.3% of retinoic users improved; tolerability was comparable. | Two patients (one per group) developed severe facial eczema and were excluded. |
| Cook 2000 Combination | Case series of more than 3100 patients | Neck, chest, arms, hands, back · Clinic peel | 70% (with 40% trichloroacetic acid) · peel gel | single peel, neutralized at a visual end point | none | The author reports smoother texture, fewer wrinkles and striae, and faded lentigines on non-facial skin. | Complications minimal (author-reported). |
| Erbağci 2000 | Randomized, single-blind, placebo-controlled, n = 58 | Not stated · Home and clinic | 15% cream; 20–70% peels · cream (home arm); peels (clinic arm) | cream once or twice daily; peels every 2 weeks · 24 weeks | placebo cream | For atrophic acne scars, repeated 70% peels gave clearly better results; the 15% home cream was better tolerated with some useful effect. | Fewer side effects with the home cream than with peels. |
| Kostarelos 2000 Industry author | Double-blind, split-site, n = 20 | Scalp · Home use | 10% · scalp lotion | twice daily · 8 weeks | placebo lotion; 0.1% betamethasone; the two combined | All cases improved with the 10% glycolic lotion; combined with betamethasone, most sites healed in about half the time of either alone. | Not reported in the abstract |
| Tsai 2000 | Double-blind, placebo-controlled, split-side, n = 12 | Lower back, forearms · Home use | 10%, pH 3.52 (≈ 6.7% free acid) · gel | once daily for 7 days, then twice daily · 3 weeks | placebo gel, pH 5.75 | Glycolic-treated skin tanned more after UVB in Asian and Caucasian subjects (and after UVA on Asian forearms); three weeks of use did not fade an existing tan. | Not reported in the abstract |
| Cotellessa 1999 Combination | Uncontrolled, n = 20 | Face · Clinic peel | 50% (with 10% kojic acid) · peel | Not stated | none (separate trichloroacetic acid group for lentigines) | Diffuse melasma cleared completely in 6 of 20 (30%) and partly in 12 of 20 (60%). | Not reported in the abstract |
| Kim 1999 | Randomized split-face, blinded evaluator, n = 26 | Face · Clinic peel | 70% · peel | every 2 weeks | Jessner's solution | Acne improved equally on both sides after three sessions; the Jessner's side peeled significantly more. | Not reported in the abstract |
| Piérard 1999 | Randomized, double-blind, vehicle-controlled | Forearms · Home use | 10% · not stated | twice daily | 0.05% retinoic acid; 2% lipohydroxy acid; vehicle | Glycolic acid showed no significant effect on aged epidermis; retinoic acid showed the most improvement. | Not reported in the abstract |
| Ash 1998 Combination | Split-area comparison, n = 10 | Abdomen, thighs · Home use | 20% (with 0.05% tretinoin or 10% L-ascorbic acid) · cream | daily · 12 weeks | the two add-ons against each other | Both glycolic combinations improved the look of white stretch marks. | Minimal irritation. |
| Kakita 1998 Combination | Randomized, double-masked, multicenter | Face · Home use | 15% or 20% lotion (with 20% azelaic acid cream) · lotion plus cream | Not stated · 24 weeks | 4% hydroquinone cream | In darker skin, the azelaic–glycolic combination faded facial hyperpigmentation as well as 4% hydroquinone. | Slightly more peeling, burning, stinging or dryness at some visits. |
| Marrero 1998 Combination | Randomized split-face, n = 18 | Face · Clinic peel | 70% (with or without 5-fluorouracil) · peel | weekly · 8 weeks; 6-month follow-up | glycolic acid alone | Actinic keratoses cleared 91.94% with 5-fluorouracil plus glycolic vs 19.67% with glycolic alone. | No significant side effects with the combination. |
| Spellman 1998 Combination | Randomized, double-masked, multicenter | Face · Home use | not stated (with 20% azelaic acid cream) · lotion plus cream | Not stated · 12 weeks | 0.025% tretinoin cream | The azelaic–glycolic combination reduced papules and inflamed lesions more than tretinoin; overall improvement was about 25% in both. | Less dryness, scaling and redness than tretinoin. |
| Thibault 1998 | Randomized, double-blind, placebo-controlled, n = 75 | Face, neck · Home use | 5% (unneutralized) · cream | daily · 3 months | placebo cream | Texture and discoloration improved significantly more than with placebo; the wrinkle trend did not reach significance. | Not reported in the abstract |
| Burns 1997 Combination | Randomized controlled pilot, n = 19 | Not stated · Home and clinic | up to 68% peels; 10% gel (with 2% hydroquinone) · peels plus home gel | six serial peels; gel twice daily and 0.05% tretinoin nightly | same home regimen without peels | In dark skin (types IV–VI) with post-inflammatory hyperpigmentation, adding peels gave a trend toward faster and greater improvement. | Normal skin also lightened in the peel group; adverse effects minimal. |
| Fartasch 1997 | Vehicle-controlled | Forearms · Home use | 4% · formulation | twice daily · 3 weeks | vehicle | Shedding was limited to the outermost layer; barrier lipids looked normal and water loss did not change. | Not reported in the abstract |
| Lawrence 1997 | Split-face comparison | Face · Clinic peel | Not stated · peel | with topical tretinoin and hydroquinone | Jessner's solution | Lightening was similar on the glycolic side (3.14) and the Jessner's side (2.96); melasma area and severity fell 63% overall. | Not reported in the abstract |
| Lim 1997 Combination | Split-face, n = 10 | Face · Home and clinic | 20–70% peels; 10% cream (with 2% hydroquinone) · peels plus home cream | peels every 3 weeks to one side; cream twice daily to both · 26 weeks | cream alone on the other side | Melasma and fine wrinkling improved on both sides; the peeled side did better but not significantly. | Not reported in the abstract |
| Wang 1997 | Uncontrolled, n = 40 | Face · Home and clinic | 35% or 50% peels; 15% home products · peels plus home products | four peels | none | Comedones, papules and pustules resolved significantly in Asian skin; scars and cysts needed repeated treatment. | Side effects in 5.6%. |
| Becker 1996 | Histology comparison, n = 2 | Face · Clinic peel | 50%; 70%, pH from below 2 to above 2 · peel solutions | 30 minutes, then neutralized · 48 hours | seven commercial solutions at different pH | Peel solutions below pH 2 caused crusting and necrosis that partially neutralized solutions above pH 2 did not; 70% caused more damage than 50%. | Crusting and necrosis below pH 2. |
| Ditre 1996 | Placebo-controlled, split-arm | Forearms · Home use | 25% (glycolic, lactic or citric acid) · lotion | Not stated · about 6 months | placebo lotion | AHA-treated forearms were about 25% thicker, with denser collagen; results are reported for the three acids together. | No inflammation seen. |
| Kopera 1996 | Within-patient comparison, n = 10 | Forearms, hands · Not stated | Not stated · commercial peeling fluid | Not stated | Q-switched ruby laser | The glycolic peeling fluid did not clear sun spots on forearms and hands; a single laser treatment cleared them by 4 weeks. | Burning, irritation and scaling with the peeling fluid. |
| Newman 1996 | Double-blind, vehicle-controlled, n = 41 | Face, forearms, hands · Clinic peel | 50% · peel | 5 minutes, once weekly · 4 weeks | vehicle on the other side | Rough texture and fine wrinkling decreased, with fewer solar keratoses and slight lightening of sun spots; biopsies showed a thinner outer layer and thicker epidermis. | Not reported in the abstract |
| Perricone 1996 | Two vehicle-controlled studies | Back · Test protocol | 12% cream (pH 4.2); 8% cleanser and lotion (pH 3.25); 50% peel (pH 2.75), pH 4.2 / 3.25 / 2.75 · cream; cleanser and lotion; peel | cream four times a day after UV; products daily · 48 hours to 3 weeks | vehicle; untreated skin | Glycolic cream reduced UV redness at 48 hours; three weeks of 8% products before UV gave an SPF of about 2.4 (authors' conclusion: photoprotective). | Not reported in the abstract |
| Piacquadio 1996 | Randomized pilot, n = 12 | Not stated · Home and clinic | 70% peels; 10% moisturizer · peel; moisturizer | monthly peels; moisturizer twice daily in half the group · 4 months | peels alone vs peels plus 10% moisturizer | 9 of 10 patients felt improved with no difference between groups; gains in fine wrinkling and pigment came mainly with the added 10% moisturizer, and no specific benefit was found for the monthly peels. | Not reported in the abstract |
| Piérard 1996 | Double-blind, split-face | Face · Home use | 6% · not stated | Not stated · 1 month | 0.05% tretinoin | Skin elasticity was significantly higher on the tretinoin side than on the glycolic side. | Not reported in the abstract |
| Stiller 1996 | Randomized, double-blind, vehicle-controlled, n = 74 | Face, forearms · Home use | 8% · cream | twice daily · 22 weeks | 8% L-lactic acid cream; vehicle cream | 76% of glycolic users improved at least one grade in facial photodamage vs 40% on vehicle; on forearms glycolic beat vehicle for overall photodamage and sallowness. | One participant (of all arms) withdrew for irritation. |
| Tse 1996 Combination | Prospective split-face, n = 13 | Face · Clinic peel | 70% (followed by 35% trichloroacetic acid) · peel | single peel · 60 days | Jessner's solution plus 35% trichloroacetic acid | The glycolic-based medium-depth peel removed actinic keratoses slightly better (response score 1.5 vs 1.0) and produced more new elastin. | Not reported in the abstract |
| Effendy 1995 | Vehicle-controlled comparison, n = 6 | Not stated · Test protocol | 12% · solution in water | 60 min once daily, 5 days a week · 2 weeks | 0.1% retinoic acid; vehicles | Glycolic acid shortened stratum corneum turnover more than retinoic acid (12.8 vs 15.8 days); water loss rose during use and hydration fell on days 11–18. | Rapid, intense redness. |
| Coleman 1994 Combination | Case series with biopsies | Face · Clinic peel | 70% (followed by 35% trichloroacetic acid) · peel | 70% applied 2 minutes, then trichloroacetic acid | none | The combined glycolic–trichloroacetic peel produced a medium-depth injury and improved pigment changes and sun damage. | Not reported in the abstract |
| Perricone 1993 | Two placebo-controlled trials, n = 35 | Beard area · Home use | Not stated · lotion | Not stated | placebo lotion | Glycolic acid lotion cut razor-bump lesions by over 60% on the treated side and allowed daily shaving with little irritation. | Not reported in the abstract |
How we built this table
Search. On 22 November 2026 we searched PubMed through NCBI's E-utilities for glycolic acid clinical trials and randomized trials in humans, for titles naming glycolic or alpha-hydroxy acids alongside trial terms, and added every PubMed study already cited on this site. That gave 392 records. We screened titles, then read every remaining abstract in full.
What is in. Any human study, of any design, in which glycolic acid was applied to skin and an effect on skin was reported, at home, in a clinic or in a test protocol. Products where glycolic acid is one named ingredient among several are included and marked Combination.
What is out. Reviews; lab, animal and cell studies; glycolic acid used in polymers, dentistry or toxicology; studies that say "AHA" without naming glycolic acid; studies where glycolic acid was in every arm and something else was being tested; and records with no abstract. 32 studies were excluded after reading their abstracts, each for a recorded reason.
How each row was written. Each abstract was read in full and only what it states was recorded; "Not stated" means the abstract does not say. The finding and side-effect columns are paraphrases, not quotes. A script then checked that every number in a row appears in its abstract. Industry author marks a study where at least one author works for a company that makes or sells skin-care or drug products, as listed in PubMed; 11 studies carry it. Free acid, where shown, is our estimate from the reported strength and pH using glycolic acid's pKa of 3.83, the same formula as our calculator; it is arithmetic, not a measurement.
Limits. We read abstracts, not full papers, so details only in the full text are missing. We did not grade study quality beyond recording the design and size each abstract reports. Many studies are small, unblinded or uncontrolled, and a positive finding in an uncontrolled study can reflect time, other treatments or chance. PubMed does not index every journal.
Reuse. The table is published under CC BY 4.0: you may copy and adapt it with a credit and a link to this page.
Frequently Asked Questions
Is there scientific evidence that glycolic acid works?
Yes, for some uses, but most of it is small. The strongest home-strength evidence is three placebo-controlled trials: a 5% cream improved facial texture and discoloration over 3 months in 75 people (Thibault 1998); an 8% cream improved facial photodamage in 76% of users against 40% on a vehicle cream over 22 weeks (Stiller 1996); and a 10% emulsion at pH 4 improved mild acne versus placebo in 120 people (Abels 2011, run by the product's maker). Most other studies are clinic peels, small, or uncontrolled. Every study is in the table above.
Does glycolic acid work for acne?
In trials, yes. A 10% emulsion at pH 4 used every evening improved mild acne more than placebo by day 45 (Abels 2011), and 40% clinic peels reduced moderate acne lesions more than placebo on the other half of the face (Kaminaka 2014). Against other treatments it mostly ties: 30% glycolic and 30% salicylic peels worked equally, though only the salicylic side stayed improved 2 months later (Kessler 2008), and a 12% cream improved acne about as often as 0.05% retinoic acid cream (Kharfi 2001). For products, see our glycolic acid cleanser roundup.
Does glycolic acid work for melasma and dark spots?
The record is mixed. Adding glycolic peels to a home cream speeded up melasma improvement in several trials (Sarkar 2002; Chaudhary 2013), but four peels added nothing to 4% hydroquinone (Hurley 2002), and 10% glycolic acid added to 4% hydroquinone improved melasma less markedly than hydroquinone alone, with the most side effects (Ibrahim 2015). A commercial glycolic peeling fluid did not clear sun spots on forearms and hands (Kopera 1996). Our dark spots guide covers the options.
Does glycolic acid work for keratosis pilaris?
The only glycolic-specific trial we found used 50% or 70% clinic peels: bumps improved by 60% at day 80, but the 9 people followed up at 5 years were no different from before treatment (Tian 2021). We found no trial of a home-strength glycolic product for keratosis pilaris. Our keratosis pilaris guide covers what the dermatology sources recommend.
Is glycolic acid better than retinol or tretinoin?
Not in the comparisons we found. Tretinoin did better on facial elasticity over a month (Piérard 1996); on aged forearm skin, retinoic acid improved the epidermis while 10% glycolic acid showed no significant effect (Piérard 1999); and tretinoin cream beat 70% glycolic peels for underarm acanthosis nigricans (Ghiasi 2024). In acne and melasma, glycolic acid tied with tretinoin creams and peels and with a 4% retinol peel (Rini 2026). We found no trial comparing a cosmetic retinol with glycolic acid at home. See our glycolic acid vs retinol comparison.
Does glycolic acid make your skin more sensitive to the sun?
Three studies say yes. Four weeks of 10% glycolic acid at pH 3.5 increased sunburn cells and lowered the UV dose needed to burn, and the effect was gone a week after stopping (Kaidbey 2003); 10% glycolic acid increased redness, DNA damage and sunburn cells after simulated sunlight while 2% salicylic acid did not (Kornhauser 2009); and treated skin tanned more after UVB (Tsai 2000). One 1996 study reported a small protective effect (Perricone 1996). Our sun exposure guide covers what to do about it.
What strength of glycolic acid did the studies use?
Home products in the trials ran from 4% to 20%, and clinic peels from 20% to 70%. The home trials with a placebo used 5%, 8% and 10% (Thibault 1998; Stiller 1996; Abels 2011). For peels, one wrinkle study found no difference between 35%, 50% and 70% (Funasaka 2001), and peel solutions below pH 2 caused crusting and necrosis that solutions above pH 2 did not (Becker 1996). No study we found compared home strengths head to head. Our percentage guide explains what each strength does.
How long did glycolic acid take to work in studies?
It depended on the problem. A 10% acne emulsion beat placebo by day 45 (Abels 2011), and a 5% acne gel showed improvement at 4 weeks (Chen 2022). For sun-damaged skin, the trials ran 3 months at 5% (Thibault 1998) and 22 weeks at 8% (Stiller 1996). Our before-and-after guide sets out realistic timelines.
Does glycolic acid work on the scalp?
One trial tested it: a 10% glycolic acid scalp lotion, used twice daily for 8 weeks, improved scalp psoriasis in all 20 patients, and combining it with a 0.1% betamethasone lotion halved the time to healing (Kostarelos 2000). The study was run with a pharmaceutical company. We found no trial of glycolic acid for dandruff or scalp buildup. See our scalp guide.
Is there evidence for glycolic acid on the underarms?
Only at clinic strength, and only for one condition. In acanthosis nigricans of the underarms and neck, tretinoin cream worked better than 70% glycolic peels on the underarms, and neither worked significantly on the neck (Ghiasi 2024). We found no trial of a home-strength glycolic product on the underarms, for darkening or for odor. Our body and underarm guide covers what the sources do say.
Are glycolic acid studies paid for by skincare companies?
Some are. We mark a study "Industry author" when PubMed lists at least one author as working for a company that makes or sells skin-care or drug products; this applies to a minority of the table, including the 10% acne emulsion trial (Abels 2011) and the trial in which a retinaldehyde cream matched a series of glycolic peels, run by researchers at Pierre Fabre, a skin-care company (Rouvrais 2018). An industry author does not make a result wrong, but it is worth knowing when you weigh it. Use the table's search to see them.
Why do so few glycolic acid studies report pH?
We cannot say why, but the gap matters. pH decides how much of the acid is free to act: at pH 4, a 10% product has about 4% free acid (Abels 2011, our estimate with glycolic acid's pKa of 3.83). Most abstracts give the percentage but not the pH, so their strength cannot be compared fairly. The same gap exists on product labels; see our statistics page and the calculator in our percentage guide.