Dapsone Gel (Aczone) for Acne: Evidence and Side Effects

Quick Summary

Yes, but the effect is small. Very large vehicle-controlled trials show topical dapsone beating its vehicle on lesion counts and clear-skin ratings by a consistent but narrow margin. It falls under the 2024 AAD guideline’s strong, moderate-certainty recommendation for topical antibiotics as a class rather than being graded on its own. The FDA label requires no G6PD screening before starting…

Close-up of clear gel with small air bubbles
Where the guideline standsStrong recommendation (2024 AAD guideline), on moderate certainty evidence
AvailabilityPrescription
How long before you judge it8–12 weeks
PregnancyDiscuss with your clinician.
Sources behind this page8, every one linked at the bottom
Dapsone Gel (Aczone) for Acne at a glance. Recommendation strength and evidence certainty are separate axes under GRADE.

What it is

Dapsone is a sulfone, first used in the 1940s against leprosy and later for dermatitis herpetiformis. Aczone is dapsone in a topical gel: the 5% twice-daily version was approved in July 2005 and the 7.5% once-daily version in February 2016. The 7.5% gel is indicated for the topical treatment of acne vulgaris in patients 9 years of age and older. What it does in acne is genuinely unsettled: the prescribing information states outright that “the mechanism of action of dapsone gel in treating acne vulgaris is not known.” It is usually described as both antibacterial and anti-inflammatory, and in practice it is reached for mainly for inflammatory papules and pustules rather than comedones.

What the evidence actually shows

Two sets of vehicle-controlled trials underpin the two strengths, and both are unusually large.

For the 5% twice-daily gel, two 12-week trials randomized roughly 700 people per arm. Success, defined as no or minimal acne, was 42% versus 32% on vehicle in one trial and 35% versus 28% in the other. Inflammatory lesions fell by 46% versus 42% and 48% versus 40%; non-inflammatory lesions by 31% versus 24% and 30% versus 21%.

For the 7.5% once-daily gel, two identically designed multicentre trials randomized 4,340 subjects. Global Acne Assessment Score success was 30% versus 21% on vehicle in both trials. Inflammatory lesions fell by 56% versus 49% and 54% versus 48%; non-inflammatory lesions by 45% versus 39% and 46% versus 41%. A pooled publication of those two trials (Thiboutot et al., J Clin Aesthet Dermatol 2016) reports 29.8% versus 21.1% success, with inflammatory lesions down 54.6% versus 48.1%.

Two things stand out. First, the vehicle does a great deal of work: roughly a fifth to a third of people on gel base alone met the success definition, so the drug’s own contribution is the gap between the columns, not the headline number. Second, the 7.5% once-daily formulation was never tested head to head against 5% twice daily, so “stronger” here means a higher concentration and a simpler routine, not demonstrated superiority.

The guideline handles dapsone inside recommendation 1.4, “For patients with acne, we recommend topical antibiotics” (strong, on moderate-certainty evidence) rather than grading it separately, and attaches the remark that topical antibiotic monotherapy is not recommended. Strength and certainty are separate axes: a strong recommendation for the class is not the same as high-certainty evidence for this drug.

How it is actually used

A pea-sized amount, once daily for the 7.5% gel and twice daily for the 5%, in a thin layer over the whole face rather than dabbed on individual spots. The label says to reassess if there is no improvement after 12 weeks, which is a fair minimum before judging it. Because the guideline warns against topical antibiotic monotherapy, dapsone is normally combined with benzoyl peroxide or a retinoid, but applying benzoyl peroxide directly on top of freshly applied dapsone causes a temporary yellow-to-orange discoloration of skin and facial hair, so people commonly separate the two, one in the morning and one at night. That workaround is a practical convention rather than a labelled instruction.

Safety and who should avoid it

Adverse events in the 7.5% trials were essentially the same as vehicle; application site dryness (1.1%) and pruritus (0.9%) were the most common. The 5% label lists no contraindications at all. Methemoglobinemia has been reported after marketing, including cases resulting in hospitalisation, and the label advises avoiding dapsone gel in people with congenital or idiopathic methemoglobinemia. Some G6PD-deficient users of the 5% gel showed laboratory changes suggestive of hemolysis, but no clinically relevant hemolysis or anemia. Human pregnancy data are absent; oral dapsone was embryocidal in animals at systemic exposures more than 250 times those from the gel. Pregnancy and breastfeeding are a conversation to have with a clinician.

Commonly repeated: you need a G6PD test before starting dapsone gel, the way you would with oral dapsone. What the evidence actually shows: no screening is required. A randomized, vehicle-controlled crossover study in 64 G6PD-deficient patients (Piette et al., Arch Dermatol 2008) found no clinical or laboratory evidence of drug-induced hemolytic anemia; the FDA label carries no screening requirement, and a 2007 labelling supplement removed the earlier blood-monitoring language; and the 2024 AAD guideline does not call for G6PD testing before topical dapsone.

Sources

Keep reading

See where this sits against every other acne treatment on the acne treatment evidence map, or check how it combines with another ingredient in the ingredient combination checker. For the guideline’s own wording on this, see the 2024 AAD guideline, recommendation by recommendation.

Not medical advice

This page reports what the published evidence says about a treatment. It is not a treatment plan and it cannot account for your skin, your other medications or your medical history. Nothing here has been reviewed by a clinician; see our editorial policy for what that means. Talk to a clinician before starting or stopping any prescription treatment.

Written by

Research, not medical advice. Every clinical claim above links to the guideline, systematic review, trial or drug label it came from, so you can check it yourself rather than take my word for it. It was researched and drafted with AI assistance by a non-clinician, and no dermatologist has reviewed it: this site does the reading, it doesn’t practise medicine. It can’t account for your circumstances either, so talk to a dermatologist or doctor about your own skin, especially before starting or stopping a prescription treatment.