Danazol (200 mg three times a day) and medroxyprogesterone acetate (MPA, 100 mg a day) were compared with placebo in the treatment of infertility of patients with endometriosis. Twenty-seven patients had medical therapy alone for 6 months, and 22 patients received it after conservative surgery. The clinical characteristics of the patients in the danazol group (n = 18), the MPA group (n = 17), and the placebo group (n = 14) were comparable to each other. The follow-up time was 30 months. The cumulative pregnancy rates, 33% in the danazol group (n = 6), 42% in the MPA group (n = 7), and 46% in the placebo group (n = 6), did not differ significantly from each other. The time to pregnancy after the start of therapy was 17.7 +/- 8.4 (standard deviation [SD]) months in the danazol group, 18.0 +/- 9.0 months in the MPA group and 10.0 +/- 5.8 months in the placebo group with no significant difference between the groups. The abortion rate was 26%, and there was no significant difference among the groups. Cox multivariant analysis did reveal ovarian endometriosis a prognostically significant negative indicator as regards fecundation in endometriosis (P less than 0.05). In summary, correction of infertility alone does not appear to be an indication for the use of danazol or MPA in the treatment of endometriosis, and ovarian endometriotic lesions but not peritoneal ones do make a worse prognosis as regards fecundation in endometriosis.