Forty patients with previously untreated epithelial ovarian cancer (OC), with FIGO stages Ic-IV, were randomly allocated to receive intravenously either carboplatin (C) 400 mg/m2 every three weeks (C arm, 20 patients) or a combination of C 350 mg/m2, ifosfamide (I) 5 g/m2 with mesna every three weeks, vincristine (V) 1.4 mg/m2 (maximum total dose 2 mg) and bleomycin (B) 30 mg every ten days (CIVB arm, 20 patients). Responding patients received 6 courses of chemotherapy and all 40 patients were evaluable for toxicity, response and survival. Clinical characteristics of patients were similar in both arms. Clinico-imaging results with chemotherapy were as follows: in C arm, clinical complete remission (cCR) 11 (55%) and partial remission (cPR) five (25%) patients for an overall response rate (ORR) 80%. For CIVB arm, cCR ten (50%) and cPR four (20%) patients for an ORR 70%. Second look operation (SLO) was performed in three of the 11 cCR patients in the C arm and six of the ten cCR patients in the CIVB arm. Pathological CR (pCR) was confirmed in two of the three cCR C arm patients and two of the six in the CIVB patients. In the first interim analysis statistically significant differences, all favoring monochemotherapy, were seen in response duration, time to progression, disease-free and overall survival. These results along with the severe myelotoxicity resulting in dose reductions in the CIVB patients led us to stop enrolling new patients and follow this population closely for long-term results. These confirmed the first observations of the superiority of single-agent carboplatin in response duration (p = 0.034), time to progression (p = 0.028), disease-free survival (p = 0.010) and overall survival (p = 0.043). Because of the above reasons we have concluded that monochemotherapy with carboplatin is to be preferred over carboplatin in combination with other drugs.