Three hundred twenty-six patients who had failed prior alkylating agents, given either as adjuvant therapy or therapy for advanced breast cancer or both, were randomly assigned to treatment with up to seven cycles of doxorubicin 75 mg/m2 or docetaxel (Taxotere, Rhône-Poulenc Rorer, Antony, France) 100 mg/m2 given every 3 weeks. The two arms of the study were well-matched for age, performance status, previous therapy, and the nature of the metastatic disease. Forty-seven percent of the docetaxel-treated patients and 49% of the doxorubicin-treated patients were defined as having disease that showed primary or secondary resistance. Overall, 48% of patients treated with docetaxel had an objective response. This was significantly greater than the 33% response rate seen in doxorubicin-treated patients (P = .008). Among those patients classified as having resistant disease, the difference was equally clear cut (response rate, 47% v 25%; P = .003). Overall median time to response was 12 weeks with docetaxel and 23 weeks with doxorubicin. Febrile neutropenia, grade 3/4 nausea and vomiting, and cardiotoxicity were significantly more common among the doxorubicin-treated patients, while diarrhea grade 3-4, skin toxicity, neurologic toxicity, fluid retention, and allergy of any grade were significantly more likely in the docetaxel-treated patients. This study demonstrates for the first time the superiority in terms of response rate of a taxoid over an anthracycline in the treatment of advanced breast cancer.