In a randomised, double-blind study the efficacy and toxicity of oral fluconazole 50 mg daily and ketoconazole 200 mg daily were compared for the treatment of oropharyngeal candidiasis in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex (ARC). 20 episodes (18 patients) were treated with fluconazole and 20 episodes (19 patients) with ketoconazole. Pretreatment clinical features and laboratory test results were similar in both groups. 17 episodes (85%) in the fluconazole group and 16 (80%) in the ketoconazole group could be evaluated. There was clinical cure at the end of therapy in all fluconazole-treated and 12 of 16 (75%) ketoconazole-treated episodes. Cultures were negative at the end of therapy in 87% of the fluconazole group and 69% of the ketoconazole group. 1 patients stopped taking fluconazole because of severe nausea. 1 of 18 fluconazole-treated and 4 of 19 ketoconazole-treated patients had transient rises in alanine or aspartate aminotransferase. Fluconazole seemed more effective than ketoconazole in the treatment of oral thrush among AIDS and ARC patients.