In 1987 a phase II study of combined high-dose platinum (carboplatin 300/mg/m2 day 1, cisplatin 50 mg/m2 days 2 and 3 q4wk) was carried out in 42 previously untreated ovarian cancer patients with residual disease. Since then, another phase II study of combined high-dose platinum and ifosfamide (1,500 mg/m2 days 1 to 3) has been carried out in 37 patients, while a third study of combined high-dose platinum and etoposide (70 mg/m2 intravenously days 1 to 5) is ongoing. Pathologic complete response (CR) and partial response (PR) rates in the first two studies were 62% in 37 evaluable patients, and 58% in 36 patients, 22% and 42% of whom were CRs, respectively. The preliminary results from the third study were: CR plus PR, 56%; CR, 24%. Hematologic toxicity was the dose-limiting factor in all three studies. Myelosuppression became substantial, but manageable, if another drug was added to the platinum combination. The percentage of patients experiencing World Health Organization grades 3 and 4 toxicity during treatment were: white blood cells 44%, 92%, and 79%; platelets 81%, 100%, and 95%, respectively, in studies I, II, and III. Nonhematologic toxicity was modest in all studies. Dose-limiting neurotoxicity occurred in 7%, 6%, and 5%; nephrotoxicity in 22%, 6%, and 11% of the patients. The percentage of patients receiving the stipulated doses of all study drugs in the sixth cycle was 31 in the first study compared with none in the second study. Combined high-dose platinum given either alone or in combination with ifosfamide or etoposide is highly active in ovarian carcinoma. However, further follow-up and a randomized trial are needed to establish the superiority of any one regimen.