Based upon results obtained in a Phase I study, we conducted a Phase II trial of high-dose CBDCA and etoposide administered via the intraperitoneal (IP) route in patients with ovarian cancer. CBDCA at a dose of 600 mg/m2 and etoposide at a dose of 400 mg/m2 were administered rapidly into the peritoneal cavity. The total dose of each agent was calculated and given daily over 3 days in amounts equal to one-third of the total dose. On day 1 of therapy, one-third of the dose was mixed in 2 liters of D5W and administered intraperitoneally as rapidly as possible. On days 2 and 3, one-third of the dose was mixed in 1 liter of D5W and administered similarly. GM-CSF was begun on day 4 as a subcutaneous injection at a dose of 500 micrograms/m2/day. A total 53 courses of treatment was administered to 18 patients; 9 of 13 patients (69%) with evaluable disease demonstrated evidence consistent with a partial response; however, the majority were response determined by a decrease in tumor marker (CA-125). One patient who had pathologic evidence of disease at second look laparotomy, but no measurable disease, was treated and shown at subsequent reexploration to have no further evidence of disease. This patient remains free of disease at 17+ months. The toxicity encountered in this trial was formidable, resulting in the removal of 78% of the patients from the study prior to completing 6 cycles of therapy.