What azelaic acid should you use for your skin?

There is no single best strength

The evidence does not name one best azelaic acid. Your target condition matters most, followed by how your skin handles irritation. For the wider picture, see our azelaic acid guide.

Rosacea: start with 15% gel or foam

For mild to moderate papulopustular rosacea, the 15% gel is the strength the FDA has labeled, applied twice daily. The label reports that mean plasma levels stayed within the range seen with vehicle alone, so systemic exposure did not rise meaningfully (FDA label for 15% gel).

In a 12-week trial of 961 patients, the 15% foam beat placebo on treatment success, 32.0% versus 23.5% (p < 0.001) (trial summary). That is a real difference, but it comes from a single trial.

A network meta-analysis of 19 RCTs with 8,208 participants ranked 20% cream highest for IGA score improvement (OR 8.54, 95% CI 2.48–29.45). The authors noted that sample sizes were limited for some agents, and that interval is wide, so read the ranking as a signal rather than a verdict (Shaheen et al. network meta-analysis).

The same analysis put 15% at the top for adverse-event odds (OR 1.95, 95% CI 1.30–2.93). Do not assume the stronger cream is automatically the harsher one, and do not assume the reverse either.

Acne and melasma: 20% cream has the most trials

For acne and melasma, the 20% cream has the deepest trial base. A systematic review of 43 RCTs found that 16 acne studies favored azelaic acid over vehicle, and 20% cream reduced more lesions than erythromycin gel (King et al. systematic review).

One review reports that 20% cream matched 0.05% tretinoin for comedonal acne, with fewer topical side effects (multiple-uses review). For melasma, seven trials found 20% cream beat vehicle on severity and global improvement (King et al.).

Other reports describe 20% cream as comparable to hydroquinone 2% or 4% (2025 comprehensive review). That is a reasonable option if hydroquinone is not suitable for you, but comparability in reports is not the same as head-to-head proof.

Where 10% fits, and why we would wait

A 10% nanocrystal hydrogel has one small acne trial behind it. Against 20% cream, given about 1 g twice daily for 8 weeks, success at week 8 was reported as 36.51% versus 30.37%, both p < 0.001 (nanocrystal hydrogel trial).

The retrieved text did not give the trial's sample size or sponsor, and we found no head-to-head 10% versus 15% data for rosacea. One trial is a reason to look, not a reason to switch. Older data also deserves caution: a 2006 systematic review of rosacea trials found the included studies were generally of modest quality (Liu et al., DARE record).

Irritation, darker skin and pregnancy

The most common side effect is mild, transient redness with scaling, itching, or burning (mechanism review). Persistent or worsening irritation is a reason to call your prescriber rather than push through.

Hypopigmentation is rare, but the label says azelaic acid has not been well studied in dark complexions, so watch for it if your skin is darker (FDA label).

The FDA label assigns Pregnancy Category B but notes there are no adequate, well-controlled topical studies in pregnant women (FDA label). A 2025 review calls it considered safe in pregnancy, but that is weaker evidence than the label, so talk it through with your clinician.