What azelaic acid is good for, and how strong the evidence is

Rosacea: where the evidence is strongest

Rosacea is the use with the most solid support. The FDA label covers topical treatment of inflammatory papules and pustules in mild to moderate rosacea, using a 15% gel applied twice daily. It says erythema without papules or pustules has not been evaluated, and it advises reassessment if there is no improvement by 12 weeks (FINACEA prescribing information).

A 2023 systematic review of 43 randomized trials, with meta-analysis of 20 rosacea trials, found improvement over vehicle at 12 weeks for inflammatory lesion counts, erythema, overall improvement, and treatment success. It also beat metronidazole 0.75% on erythema, overall improvement, and inflammatory lesions (systematic review of azelaic acid across four skin conditions).

The most recent head-to-head comparison is a 2024 network meta-analysis of 19 trials with 8,208 participants. It ranked the 20% cream highest on clinician-rated success in moderate-to-severe papulopustular rosacea, while the 15% strength carried the highest adverse-event risk. The authors note that some comparator groups were small and that many clinically important outcomes were overlooked (network meta-analysis of rosacea treatments).

An older 2006 review of five trials with 873 patients found azelaic acid reduced inflammatory lesions and erythema. Pooling was not possible because standard deviation data were missing for four of the five studies, and it showed no significant effect on telangiectasia (Liu et al., 2006 systematic review). For the broader picture across uses, see our azelaic acid topic page.

Acne: a reasonable second-line option

Azelaic acid is commonly used at 15% to 20% for acne, and a mechanistic review describes it as second-line. We did not find a primary systematic review pooling acne trials, so the comparisons here come from narrative sources.

A 2025 review reports lesion reductions comparable to topical retinoids and adapalene, with better tolerability in some studies. Treat that as directional, not settled. DermNet says it is probably similarly effective to tretinoin with milder side effects, but it may work less well than benzoyl peroxide on participants' own global self-assessments (DermNet overview).

Acne results take time. DermNet notes that improvement typically appears around four weeks, with the best results after about six months. The 2025 comprehensive review is the source for the retinoid comparison, so read it as a narrative summary rather than a pooled analysis.

Melasma and dark spots: promising, less settled

This is the use people search for most, and the evidence is the weakest of the three. A 2025 review reports that 20% cream matched hydroquinone for melasma and other dyschromias, with a favorable safety profile (2025 comprehensive review). We could not verify the primary numbers behind that claim, so treat it as a lead rather than a verdict.

The mechanism explains the interest in darker skin. Azelaic acid inhibits tyrosinase and does not appear to affect normal melanocytes, and both a mechanistic review and DermNet tie its profile to use in darker skin. The catch is that hypopigmentation is a recognized side effect, and it is more visible in darker skin tones.

Safety: mostly mild, with a few things to watch

Mild, transient local irritation is the most common complaint across uses (2025 comprehensive review). FDA pharmacokinetic data show the 15% gel does not raise plasma azelaic acid beyond the range seen from diet and the body's own metabolism (FINACEA prescribing information).

• Early irritation is common and usually mild, so give your skin time to adjust.
• Watch for lightening in darker skin tones, which is a recognized adverse effect.
• If rosacea shows no improvement by 12 weeks, the FDA label advises reassessment.
• Pregnancy and long-term safety data were not in the sources we retrieved, so ask a clinician first if you are pregnant or planning to be.

Where the evidence runs thin

Several gaps matter. We found no primary systematic review pooling acne trials. Pooled effect sizes, sample sizes, and GRADE certainty ratings for melasma were not retrievable. We also could not confirm data for a 30% concentration mentioned in one review.