OBJECTIVE: To compare the survival between intraperitoneal cisplatin-based chemotherapy (IPCT) and intravenous cisplatin-based chemotherapy (IVCT) in stage III epithelial ovarian cancer with minimal residual disease (<1 cm) after primary debulking surgery. METHOD: One hundred and thirty-two patients with stage III epithelial ovarian cancer after optimal primary debulking surgery with minimal residual disease between April 1990 and March 1995 were entered into a randomized clinical trial in which IPCT or IVCT was administered at 3-week intervals. Patients in the IPCT arm received cisplatin-based (100 mg/m(2)) intraperitoneal chemotherapy. Patients in the IVCT arm received cisplatin-based (50 mg/m(2)) intravenous chemotherapy. The tumor response was assessed every 3 months. The hematological toxicity using the South West Oncology Group (SWOG) toxicity criteria was assessed. Catheter complications associated with intraperitoneal chemotherapy were also analyzed. RESULT: The estimated median survival in the IPCT group was 43 months (95% confidence interval, 34-54) and IVCT group was 48 months (95% confidence interval, 37-59). The hazard ratio of death was not statistically significant between IPCT and IVCT (hazard ratio, 1.13; 95% CI, 0.69-1.86; P=0.317). The frequencies of hematological toxic effects were significantly lower in the IPCT group than in the IVCT group. CONCLUSION: Intravenous and intraperitoneal chemotherapy are associated with equivalent survival in patients with minimal residual stage III epithelial ovarian cancer after optimal cytoreductive surgery.