50 patients with advanced breast cancer were treated with the combination of Mitoxantrone 10mg/m2 IV day 2, 5-Fluorouracil 400mg/m2 IV and Cyclophosphamide 300mg/m2 IV day 3, 4, 5, 6 of each monthly cycle. 49 patients are evaluable for toxicity and 47 for efficacy after three months of treatment. Hematologic toxicity was substantial and dose-limiting, with one toxic death early in the trial. Other toxicities were moderate and manageable in this short-term study. The response rate after three cycles was 53% +/- 14% with 4 complete remissions, 21 partial remissions, 16 stable disease and 6 progressions. Using the fixed response rate hypothesis of Gehan generalised by Lee and Wesley, with an expected response rate of 60% consistent with the reported response rate of advanced breast cancer to Adriamycin containing regimens, we conclude that the combination studied is not less efficient for the induction of remissions in advanced breast cancer than comparable combinations with Adriamycin. As there is now substantial experimental and clinical evidence of reduced toxicity, mainly on the cardiac muscle, of Mitoxantrone as compared to Adriamycin, we feel that the routine substitution of the latter by the former in chemotherapy for advanced breast cancer is justifiable.