From May 1978 until November 1980, 169 previously untreated patients with advanced epithelial ovarian cancer were entered into a prospective randomized clinical trial comparing the combination of hexamethylmelamine, Adriamycin, and cyclophosphamide (HAC) to a combination of melphalan and cis-platinum. Eleven patients were excluded from analysis and another 5 patients were excluded from response analysis. Of 153 patients evaluable for response, there were 47, or 30.7%, complete responders (all determined surgically), 6 partial responders, and 100 nonresponders. The response rate for the HAC group was 31% and for the melphalan-platinum group was 37.8%. The overall response rate was 34.6%. Residual tumor diameter (less than or greater than 2 cm) exerted a statistically significant effect on response--47.8 vs 24.4%. Of the 47 complete responders, 7, or 14.9%, have relapsed, with the median duration of remission of 44+ months. Of the 158 patients evaluable for survival, 90 patients have died, with a median survival time of 27.9 months (HAC = 26.4 months, melphalan-platinum = 29.6 months). Age, FIGO stage, histologic grade, and residual disease all exerted a significant effect on survival time. Second-line therapy in the treatment failures was of no benefit. Hematologic toxicity was greater in the melphalan-platinum group. Gastrointestinal toxicity was severe in both groups. Other toxicities were minor and infrequent.