Forty-four patients with measurable metastatic breast cancer have been entered in a randomized study comparing mitoxantrone to doxorubicin as a component of front-line combination chemotherapy. Patients were stratified according to whether or not they had previously received adjuvant chemotherapy. Initial doses of cyclophosphamide and 5-fluorouracil were 500 mg/m2 for both regimens, with either mitoxantrone, 10 mg/m2, or doxorubicin, 50 mg/m2. All drugs were given on day 1 only, with treatments repeated every 3 weeks. Doses were adjusted according to blood count nadirs. Responses have been observed in both of the treatment groups, though it is too early to determine the relative efficacy of the two regimens. Toxicity was comparable on the two treatment regimens except that far less alopecia and stomatitis were associated with the mitoxantrone therapy. No congestive heart failure has been seen. The combination of cyclophosphamide-mitoxantrone-5-fluorouracil is reasonably well tolerated, with myelosuppression being dose limiting. In both treatment groups of this study, reliance on the leucocyte count, rather than the granulocyte count, as a basis for dose alteration or treatment delay would lead to excessive dose reductions, many fewer dose escalations, and much more treatment delay.