To determine the relative efficacy and toxicity of stibogluconate and ketoconazole for the treatment of cutaneous leishmaniasis, a comparative trial was conducted in which 120 Guatemalan men with parasitologically proven cutaneous leishmaniasis were randomly divided into three treatment groups: sodium stibogluconate (20 mg of antimony per kilogram per day intravenously for 20 days), ketoconazole (600 mg per day orally for 28 days), and placebo. Treatment outcome was influenced by species. Among patients infected with Leishmania braziliensis, 24 (96%) of 25 in the stibogluconate group but only 7 (30%) of 23 in the ketoconazole group responded. Among Leishmania mexicana-infected patients, only 4 (57%) of 7 in the stibogluconate group but 8 (89%) of 9 in the ketoconazole group responded. These differences emphasize the importance of speciation in the treatment of leishmaniasis.