Sixty-six hirsute women were randomized and treated with 1) flutamide (n = 15), 250 mg/day; 2) finasteride (n = 15), 5 mg/day; 3) ketoconazole (n = 16), 300 mg/day; and 4) ethinyl estradiol (EE)-cyproterone acetate (CPA; n = 20), 0.01 mg EE/day for the first week, 0.02 mg EE/day for the second week, and 0.01 mg EE/day for the third week, followed by a pause of 7 days, then 12.5 mg CPA/day added during the first 10 days of every month for 12 months. Hirsutism was evaluated by the Ferriman-Gallwey score, and hair diameter and hair growth rate were determined by a special image analysis processor in basal conditions and after 90, 180, 270, and 360 days of treatment. All treatments produced a significant decrease in the hirsutism score, hair diameter, and daily hair growth rate: flutamide, -55 +/- 13%, -21 +/- 14%, and -37 +/- 18%; finasteride, -44 +/- 13%, -16 +/- 12%, and -27 +/- 14%; ketoconazole, -53 +/- 18%, -14 +/- 12%, and -30 +/- 21%; and EE-CPA, -60 +/- 18%, -20 +/- 11%, and -28 +/- 21%. Some differences existed among treatments with regard to effectiveness; EE-CPA and flutamide seem to be the most efficacious in improving hirsutism. For the hirsutism score, a greater decrease was seen with EE-CPA (-60 +/- 18%) than with finasteride (-44 +/- 13%; P < 0.01) and a greater decrease was seen with flutamide (-58 +/- 18%) than with finasteride (-44 +/- 13%; P < 0.05). Flutamide is the fastest in decreasing hair diameter; EE-CPA is the fastest in slowing down hair growth, even though at the end of the treatment there was a significant difference between flutamide and finasteride only (-41 +/- 18% vs. -27 +/- 14%; P < 0.05). Flutamide, ketoconazole, and EE-CPA induced a significant decrease in total and free testosterone, 5alpha-dihydrotestosterone, dehydroepiandrosterone, dehydroepiandrosterone sulfate, and androstenedione plasma levels. During the EE-CPA treatment, gonadotropins were suppressed, and the sex hormone-binding globulin level increased. Finasteride induced a decrease in dehydroepiandrosterone sulfate and 5alpha-dihydrotestosterone and an increase in testosterone levels. Very few side-effects were observed during treatment with low doses of flutamide, EE-CPA, and particularly finasteride. Flutamide induced a decrease whereas EE-CPA induced an increase in triglycerides and cholesterol, showing higher values within the normal range. Ketoconazole induced several side-effects and complications, and several people dropped out of the study. Despite different modalities of action and significantly different effects on androgen levels, low doses of flutamide, finasteride, and EE-CPA constitute very satisfactory alternative therapeutic regimens in the treatment of hirsutism.