Only a handful of places left for our September 2026 cohort. Applications close soonApply now →

MSTA
Your future
starts here.

Azelaic acid

Also known as: azelaic, azelaic acid cream, azelaic acid gel, Finacea

Azelaic acid is a topical dicarboxylic acid used in specific medicinal formulations for acne and papulopustular rosacea, with narrower off-label evidence for melasma and post-inflammatory hyperpigmentation. Results depend on the finished product and indication; direct human pregnancy and breastfeeding outcome data are sparse.

Evidence status

Moderate

Strongest for specific prescription formulations in papulopustular rosacea, moderate for acne. Melasma and PIH evidence is limited and formulation-specific. Low systemic exposure explains cautious clinical selection in pregnancy, but direct human outcome data are far too sparse for a proven-safe claim.

Definition#

Azelaic acid is a dicarboxylic acid — not an alpha or beta hydroxy acid, despite where it usually gets filed. In the UK it is used in specific licensed medicinal formulations for acne and papulopustular rosacea, with narrower off-label evidence in melasma and post-inflammatory hyperpigmentation.

The single most important thing to hold about it: the evidence belongs to particular finished products at particular strengths in particular indications. It does not travel to every cosmetic that lists azelaic acid on the back.[16]

UK status: this is a medicine#

Finacea 15% gel and Skinoren 20% cream are both prescription-only medicines. Finacea is licensed for facial papulopustular acne and papulopustular rosacea; Skinoren for acne vulgaris.[1, 2]

Lower-strength azelaic acid does appear in cosmetics. Where the boundary sits is less tidy than commonly stated: MHRA borderline guidance sets no fixed 10%, 15% or 20% cutoff. Classification turns on the claims made, presentation, intended purpose and mode of action.[4]

So a 10% cosmetic serum is not automatically lawful because of its percentage, and it certainly does not inherit the trial evidence built on 15% and 20% medicines.

Acne#

NICE includes topical azelaic acid at 15% or 20% in selected acne treatment and maintenance pathways, so it is a legitimate option.

But here is the finding the trade almost never repeats. Cochrane compared it against benzoyl peroxide and concluded azelaic acid probably leads to a worse treatment response.[1, 2, 3, 5]

That is a materially different statement from “not superior”. It is worth saying plainly, because azelaic acid is often positioned as the gentle-but-equally-good alternative, and on the acne evidence it is not equally good — it is a reasonable option chosen for other reasons, usually tolerability or pregnancy context.

Rosacea#

This is where azelaic acid is strongest. The Cochrane vehicle-controlled comparison covered 1,179 participants and carried high GRADE certainty, and BAD guidelines give a strong recommendation for azelaic acid, ivermectin or metronidazole as first-line topicals.[1, 6, 7]

Read the phenotype carefully, though. The evidence and the licence both concern papulopustularrosacea — the papules and pustules. They do not establish benefit for flushing, persistent erythema alone, telangiectasia, phymatous change or ocular rosacea.

A client with a red face is not automatically a client for azelaic acid, and that distinction is where a lot of disappointing outcomes come from.

Pigmentation#

Both pigment uses are off-label.

Melasma

The evidence is old, heterogeneous and generally poor quality. 20% cream outperformed 2% hydroquinone in one trial (340 randomised, 300 completing) but was not superior to 4% hydroquinone in another (329 enrolled, 243 analysed). Long-term response remained uncertain.[8]

Worth knowing who funded that literature: the pivotal 4%-hydroquinone trial was conducted in association with Schering AG, and the review found 11 of 20 melasma trials were industry-sponsored.

Post-inflammatory hyperpigmentation

Small, formulation-specific trials suggest 15–20% may improve acne-associated PIH. The limits matter: one 12-week vehicle-controlled study randomised 72 and analysed 60, predominantly phototype III; the comparative study against tranexamic acid randomised 82 with 60 completing, had no untreated control, and used sunscreen throughout.[9, 10]

None of that transfers to an undiagnosed dark mark of unknown cause.

The pregnancy question#

Azelaic acid is the active everyone reaches for when a client is pregnant and their retinoid stops. It is a defensible choice. It is not the settled matter it is usually presented as, and the difference is worth understanding before you say it out loud to a client.

What actually underpins it:

  • Low systemic exposure. The widely quoted 3.6% absorption figure comes from urinary recovery in six healthy men after a single large-area application of one 20% cream. It is not a universal rate.[1, 12]
  • Animal data.Oral animal studies found no teratogenicity, but did report embryotoxicity alongside maternal toxicity at high exposure, with no-effect levels around 3–32 times the maximum recommended human dose by body surface area.
  • Human data. The original clinical programme documented two pregnancy exposures. Both outcomes were reported as apparently healthy, with full term explicitly documented for one.[1, 2, 11, 12]i

Two exposures is not a safety dataset. The honest formulation is that azelaic acid is chosen in pregnancy because exposure is low and the animal signal is reassuring, not because human outcomes have been studied and found safe.

Breastfeeding is thinner still: no measured human-milk or breastfed-infant outcome study exists. The low-risk language in the reference sources is an exposure-based inference. Advice is to avoid breast and nipple application and infant contact.[1, 2, 13]i

And in any case: it is a prescription medicine. The pregnancy decision belongs to the prescriber, not the treatment room.

Mechanism#

Azelaic acid is described as antibacterial, anti-keratinising, anti-inflammatory and a tyrosinase inhibitor. Those proposed actions draw on a mixture of clinical, purified-enzyme, cultured-cell and animal evidence, and no single molecular story fully explains the clinical outcomes.[1, 2, 14, 15]i

The tyrosinase claim deserves particular scepticism. In the purified-enzyme work usually cited, direct inhibition was weak, and observed only at cytotoxic concentrations. A concentration that kills the cell is not a plausible mechanism for a cosmetic effect on living skin.[15]

The clinical benefit in pigmentation is better established than any proposed explanation for it. That is an acceptable position to hold, and more honest than picking a mechanism because it sounds tidy.

Irritation and tolerability#

Burning, itching, pain, redness, dryness and exfoliation are all recognised, and they are frequently treatment-limiting — in one comparative trial, 11 participants discontinued azelaic acid because of side effects.[1, 2, 9, 10]

The product information also notes it penetrates damaged skin faster than intact skin, which is exactly the wrong direction for a client who has arrived with a compromised barrier.

Two claims it does notsupport: one 12-week trial of a 15% gel found no between-group advantage over vehicle for TEWL, hydration or sebum. So “it repairs the barrier” and “it controls oil” are both unestablished.[9]

And there is no primary evidence that a purging phase demonstrates it is working.[no source found]

In professional practice#

  • Know when you are talking about a medicine. 15% and 20% are POM. Refer rather than recommend.
  • Match the rosacea phenotype. Papulopustular, not every red face.
  • Don’t position it as equal to benzoyl peroxide in acne — the evidence says otherwise.
  • Be precise on pregnancy.“Commonly chosen because exposure is low” is defensible. “Proven safe” is not.
  • Expect irritation, and don’t start it on an already-compromised barrier.

What remains uncertain#

  • Whether cosmetic-strength azelaic acid does anything comparable to the medicines.
  • Human pregnancy and breastfeeding outcomes — the data barely exist.
  • Which mechanism actually accounts for the pigment effect.
  • Long-term melasma response, which the review left unresolved.
  • How much of the PIH result is azelaic acid and how much is the sunscreen used alongside it.

Common misconceptions#

“Azelaic acid is the pregnancy-safe active.”

It is the commonly chosen one, on the basis of low absorption and animal data. The original licensing programme documented two exposed pregnancies — but that is the programme, not the literature, and this page treated it as the whole picture until August 2026. What is still true is that adequate product-specific pregnancy safety has not been established.[1, 2, 11, 12]i

“It’s as good as benzoyl peroxide, just gentler.”

Cochrane found it probably produces a worse treatment response than benzoyl peroxide.[1, 2, 3, 5]

“It treats rosacea.”

It treats papulopustular rosacea. Flushing, telangiectasia and phymatous change are not the same target.[1, 6, 7]

“It works by inhibiting tyrosinase.”

Weakly, and only at cytotoxic concentrations in a test tube. That is not a usable explanation.[15]

“It’s an AHA.”

It is a dicarboxylic acid. Different chemistry, different behaviour, no exfoliant equivalence.

Frequently asked questions#

Can I recommend azelaic acid to a client?

You can discuss it, and refer. The strengths with evidence are prescription-only in the UK, so supplying or directing use of those is a prescriber’s decision.[1, 2]

Is a 10% cosmetic serum equivalent?

No — but not for the reason this page used to give. It said there was no trial evidence below 15%. There is: a double-blind randomised trial compared a 10% azelaic-acid nanocrystal hydrogel— an engineered delivery system built to compensate for the lower concentration — with a 20% cream over eight weeks, and it performed comparably.[16]

Which makes the real point sharper. The distinction is the finished formulation, not the percentage on the front. A specific nanocrystal hydrogel at 10% is not the same object as a retail 10% serum, and neither is interchangeable with a licensed 15–20% medicine.

Is it safe while breastfeeding?

No measured human-milk or infant outcome study exists. The low-risk position is inferred from exposure. Avoid breast and nipple application and infant contact, and refer the decision.[1, 2, 13]i

Will it help my client’s dark marks?

Possibly, if they are acne-associated PIH — that is what was studied, in small trials with real limitations. For pigmentation of unclear cause, the answer is diagnosis first.[9, 10]

Does the stinging mean it’s working?

No. Irritation is common and treatment-limiting, and no primary evidence links a purging phase to efficacy.[no source found]

References#

Each source is graded by evidence tier. Tier 4 material (manufacturer documents, expert consensus, practitioner experience) is useful for protocol and context, and is never presented as equivalent to independent clinical evidence.

  1. LEO Laboratories Limited. Finacea 15% Gel: Summary of Product Characteristics. Electronic Medicines Compendium. Text revised 13 July 2023.Tier 1Supports: UK prescription status, 15% formulation, acne and papulopustular-rosacea indications, adverse effects, and the note that absorption is faster through damaged than intact skin. Oral animal studies found no teratogenicity but reported embryotoxicity with maternal toxicity at high exposure; animal NOAELs correspond to roughly 3–32× the maximum recommended human dose by body surface area.
  2. LEO Laboratories Limited. Skinoren 20% Cream: Summary of Product Characteristics. Electronic Medicines Compendium. Last updated 12 September 2023.Tier 1Supports: UK prescription status, 20% formulation, acne-only indication, common local reactions, uncommon depigmentation, pregnancy caution, and modelled — not measured — lactation exposure.
  3. National Institute for Health and Care Excellence. Acne vulgaris: management. NICE guideline NG198.Tier 1Supports: Defines topical azelaic acid as 15% or 20% and includes it in selected acne treatment and maintenance pathways for England.
  4. Medicines and Healthcare products Regulatory Agency. Borderline products: how to tell if your product is a medicine.Tier 1Supports: UK classification turns on claims, presentation, intended use and pharmacological properties. The guidance states no fixed azelaic-acid concentration boundary.
  5. Liu H, Yu H, Xia J, et al. Topical azelaic acid, salicylic acid, nicotinamide, sulphur, zinc and fruit acid (alpha-hydroxy acid) for acne. Cochrane Database of Systematic Reviews. 2020;(5):CD011368.Tier 1Supports: Compared with benzoyl peroxide, azelaic acid PROBABLY LEADS TO A WORSE TREATMENT RESPONSE — not merely an absence of superiority. Certainty for the acne comparisons was moderate to low.Funding / interest: The review notes extensive pharmaceutical support across the underlying trial evidence.
  6. van Zuuren EJ, Fedorowicz Z, Carter B, van der Linden MMD, Charland L. Interventions for rosacea. Cochrane Database of Systematic Reviews. 2015;(4):CD003262.Tier 1Supports: Vehicle-controlled rosacea efficacy across 1,179 participants, predominantly papulopustular. The vehicle comparison carried HIGH GRADE certainty. Response duration was incompletely reported.Funding / interest: 66 of 106 included studies stated funding, mostly pharmaceutical. This describes the underlying evidence base, not authorship of the review.
  7. Hampton PJ, Berth-Jones J, Duarte Williamson CE, et al. British Association of Dermatologists guidelines for the management of people with rosacea 2021. British Journal of Dermatology. 2021;185(4):725–735.Tier 1Supports: Strong recommendation for azelaic acid, ivermectin or metronidazole as first-line topical options in papulopustular rosacea, with irritancy discussed.Funding / interest: No direct funding. Several authors disclosed relationships with pharmaceutical and skincare companies.
  8. Rajaratnam R, Halpern J, Salim A, Emmett C. Interventions for melasma. Cochrane Database of Systematic Reviews. 2010;(7):CD003583.Tier 1Supports: Old, heterogeneous, generally poor-quality melasma trials. 20% cream outperformed 2% hydroquinone (340 randomised, 300 completing) but NOT 4% hydroquinone (329 enrolled, 243 analysed). Long-term response remained uncertain.Funding / interest: The pivotal 4%-hydroquinone trial (Baliña 1991) was conducted in association with Schering AG — the same manufacturer disclosed on the absorption study at source 12. The review found 11 of 20 melasma trials were industry-sponsored.
  9. Shucheng H, Zhou X, Du D, Li J, Yu C, Jiang X. Effects of 15% azelaic acid gel in the management of post-inflammatory erythema and post-inflammatory hyperpigmentation in acne vulgaris. Dermatology and Therapy. 2024;14(5):1293–1314.Tier 2Supports: Twelve-week vehicle-controlled study, 72 randomised and 60 analysed, predominantly phototype III. Notable attrition, retrospective registration, early local effects, and NO between-group advantage for TEWL, hydration or sebum.Funding / interest: Funded by the National Natural Science Foundation of China; formulations supplied by Kelun Pharma. Authors declared no conflicts.
  10. Sobhan M, Alizadeh P, Sheykhi N. A comparative study of 20% azelaic acid cream versus 5% tranexamic acid solution for postinflammatory hyperpigmentation in patients with acne vulgaris: a single-blinded randomized clinical trial. Journal of Research in Medical Sciences. 2023;28:18.Tier 2Supports: Active-comparator acne-PIH study, 82 randomised and 60 completing, with 11 azelaic-acid discontinuations for side effects, sunscreen co-intervention throughout, and no untreated control arm.
  11. US Food and Drug Administration, Center for Drug Evaluation and Research. Finacea (azelaic acid) gel 15%, NDA 21-470: Medical Review, Part 2. 2002.Tier 1Supports: Two pregnancies exposed to active 15% gel across the original clinical programme. Both outcomes reported as apparently healthy, with full term explicitly documented for one. Far too little to establish fetal safety.
  12. Täuber U, Weiss C, Matthes H. Percutaneous absorption of azelaic acid in humans. Experimental Dermatology. 1992;1(4):176–179.Tier 3Supports: The widely quoted 3.6% absorption estimate derives from urinary recovery in six healthy men after a single large-area application of one 20% cream.Funding / interest: Study affiliation was the Institute of Pharmacokinetics, Schering AG — the manufacturer associated with the product studied.
  13. National Institute of Child Health and Human Development. Azelaic acid. Drugs and Lactation Database (LactMed). Updated 15 March 2026.Tier 4Supports: States that topical use has not been studied in breastfeeding. The low-risk assessment is inferred from exposure, with no breast or nipple application and no infant contact advised.
  14. Coda AB, Hata T, Miller J, et al. Cathelicidin, kallikrein 5, and serine protease activity is inhibited during treatment of rosacea with azelaic acid 15% gel. Journal of the American Academy of Dermatology. 2013;69(4):570–577.Tier 3Supports: Cultured human keratinocyte, mouse-skin and open-label human findings on the rosacea inflammatory pathway. 60 entered, 55 completed, 49 analysed. Contains NO pigment data and does not support a melanogenesis claim.Funding / interest: Supported in part by an investigator-initiated Bayer/Intendis grant; several authors disclosed consultancy or research relationships.
  15. Schallreuter KU, Wood JW. A possible mechanism of action for azelaic acid in the human epidermis. Archives of Dermatological Research. 1990;282(3):168–171.In vitroTier 3Supports: Purified-enzyme and in-vitro work finding only weak direct tyrosinase inhibition, and at CYTOTOXIC concentrations. Not a human pigmentation outcome study.
  16. Tomić I, Miočić S, Pepić I, Šimić D, Filipović-Grčić J. Efficacy and safety of azelaic acid nanocrystal-loaded in situ hydrogel in the treatment of acne vulgaris. Pharmaceutics. 2021;13(5):567.Tier 2Supports: REFUTES THE CLAIM THAT NOTHING BELOW 15% HAS BEEN TRIALLED. A double-blind randomised trial comparing a 10% azelaic-acid NANOCRYSTAL in-situ hydrogel with a 20% azelaic-acid cream, about 1 g twice daily for eight weeks. At week eight the hydrogel reached a 36.51% success rate against 30.37% for the cream, with total inflammatory lesions down 39.15% against 33.76%. SCOPE LIMIT THAT MATTERS MORE THAN THE RESULT: this is one experimental nanocrystal delivery system, engineered specifically to compensate for the lower concentration. It says nothing about an ordinary retail 10% serum.Funding / interest: One author, Sandra Miočić, is affiliated with R&D, PLIVA Croatia Ltd., a Teva Group member.

Continue learning with MSTA#

Skinipedia is written by the educators at MSTA, the medic-led skincare training academy in Liverpool.