In a multicenter prospective randomized trial, the efficacy and toxicity of low-dose (400 mg/d) and high-dose (800 mg/d) oral ketoconazole were compared in 80 patients with blastomycosis and in 54 with histoplasmosis. Among 65 patients with blastomycosis treated for 6 months or more, high-dose treatment was more effective (100% success rate compared with 79%; p = 0.001). Among 19 patients with chronic cavitary histoplasmosis treated for 6 months or more, both regimens were equally effective (overall success rate, 84%). In 20 patients with localized or disseminated histoplasmosis treated for 6 months or more, low-dose treatment was more effective (100% success rate compared with 57%; p = 0.03). The success rate for all patients with histoplasmosis treated for 6 months or more was 85%. Adverse effects occurred in 81 of 134 patients (60%) and were commoner with the high-dose regimen. Ketoconazole is effective for immunocompetent patients with non-life-threatening, nonmeningeal forms of blastomycosis and histoplasmosis. Because of the higher frequency of side effects associated with the high dose, ketoconazole therapy should be initiated with the low dose.