One hundred thirty-three evaluable patients with advanced breast cancer entered a randomized trial comparing epirubicin 60 mg/m2 with a combination of epirubicin 45 mg/m2 and vindesine 3 mg/m2 day 1 and 8 every 4 weeks. In all 10 premenopausal women an oophorectomy was performed. Seventy-five patients had previously received cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) for advanced disease and 68 had received adjuvant chemotherapy (cyclophosphamide or CMF). Among evaluable patients (72 in the epirubicin group and 61 in the epirubicin + vindesine group) response rates were as follows: complete response--seven versus six; partial response--31 versus 22; no change--16 versus 17 (p greater than 0.40). Median time to disease progression was 6 months in both groups and median survival times were identical (12 months). Thrombocytopenia was less frequent in the epirubicin + vindesine group (p less than 0.01). In the epirubicin + vindesine group, mild to moderate peripheral neuropathy was observed in 40% of the patients. Congestive heart failure developed in one patient with a cumulative dose of epirubicin less than 1000 mg/m2 and in 7 of 15 patients who had greater than 1000 mg/m2. Four died of this cause. In conclusion, epirubicin is effective as a single agent for advanced breast cancer. The combination with vindesine does not increase its efficacy.