Thirty-seven patients with hormonally refractory prostatic carcinoma entered a randomized trial comparing doxorubicin and doxorubicin plus cisplatin. All patients had failed prior hormonal treatment. Mean Karnofsky performance status (76% doxorubicin versus 75% combination), percent of patients with prior palliative irradiation (40% doxorubicin versus 35% combination), and hemoglobin levels of less than or equal to 12 g/dl (30% doxorubicin versus 24% combination) were roughly equivalent in the two treatment groups. More patients treated with doxorubicin than the combination treatment had an elevated acid phosphatase level at study entry (90% versus 65%). Measurable bidimensional tumors were present in 13 patients in 16 sites in the doxorubicin arm and in 10 patients in 11 sites in the combination arm. Partial responses were seen in 1 of 13 patients in the doxorubicin arm and 2 (20%) of 10 patients in the combination arm. Improvement in Karnofsky performance status of 20% or greater was rarely observed with either treatment (7% doxorubicin versus 8% combination). Acid phosphatase levels normalized or improved by 50% in 39% of patients who received doxorubicin and 27% of patients who received the combination. The overall response rate by National Prostatic Cancer Project Criteria was 53% for doxorubicin and 59% for doxorubicin plus cisplatin. Myelotoxicity and gastrointestinal toxicity were severe, particularly in the combination arm, and required discontinuation of treatment in some patients who responded to treatment. Moderate renal dysfunction (creatinine value 2.0-3.0 mg/dl) occurred only in the combination arm at an incidence of 23%. Time to progression and survival were similar for the two treatment groups. In this small group of 37 patients, the combination of cisplatin and doxorubicin showed no improvement over doxorubicin alone in response, response duration, or survival, and was difficult to administer in this patient population.