The management of metastatic prostate cancer that has relapsed after initial hormonal manipulation remains a major problem, with the majority of patients dying within 12 months. Their clinical course is frequently characterized by progressive debilitation, pain, and other tumor-related symptoms. A phase II, non-randomized multicenter clinical trial was carried out in Australia in 1985-1986 to assess the efficacy and toxicity of mitoxantrone. Substantial anticancer activity was shown against hormone-refractory prostate cancer, indicated by reduction in tumor-related symptoms, improvement in quality of life indices, and a median survival of 10 months in patients with a heavy tumor burden. Although it is not possible to equate this nonrandomized series more fully with current experience since routine prostate-specific antigen measurement was not performed, the median survival of 10 months was equivalent to or better than the survival times reported from most other institutional reports of the time. Even more importantly, however, major improvements were noted in such subjective indices as reduction in pain, weight gain, and performance status. Toxicity was also acceptable, with the major side effect being asymptomatic myelosuppression.