A randomized study to compare the efficacy of combination chemotherapy (cisplatin, doxorubicin, cyclophosphamide: PACe) with chlorambucil (CB) in International Federation of Gynecology and Obstetrics (FIGO) stage III and IV ovarian carcinoma was conducted between May 1979 and October 1983. Patients failing initial CB were subsequently eligible for treatment with PACe. Eighty-nine patients were randomized and 85 were eligible for analysis; as of date, 72 of these patients have died. The majority of patients in this study had bulky residual disease after their initial laparotomy (76%). Complete response (CR) was documented by a second laparotomy after five cycles of combination therapy or 6 to 12 months alkylating agent therapy. The overall response rate (CR plus partial response [PR]) for the combination (PACe, 68%) was significantly higher (P = .0004) than that for the chlorambucil (CB, 26%). However, the median survival was not improved (PACe, 13 months; CB, 11 months) and the survival curves were not significantly different (log rank test P = .25). The results of this study are comparable to preliminary data reported from other similar randomized studies. PACe, as administered in this study, is not indicated as routine therapy in patients with bulky residual ovarian carcinoma.