To explore the clinical applicability of the Goldie and Coldman hypothesis, we treated 28 patients with metastatic breast cancer with alternating non-cross-resistant chemotherapy. The patients received cyclophosphamide, 600 mg/m2, 5-fluorouracil, 600 mg/m2, methotrexate, 40 mg/m2, alternated every three weeks with adriamycin, 60 mg/m2, and mitomycin C, 10 mg/m2. Only one patient had previously received palliative chemotherapy. Six patients had received adjuvant CMF, and 17 patients had been pretreated with endocrine therapy (13 for advanced disease, 4 as adjuvant). Fourteen patients had bone involvement, and 10 had visceral metastases. A mean of 12 cycles was given to 24 evaluable patients. The objective response rate was 67%: 11 patients (46%) achieved complete and 5 (21%) partial remission. Response rate in soft tissues was 83.3%, in bone 50%, in liver 100%, and in lung 80%. The median duration of response was 14 months, with 7 patients still in remission. No life-threatening toxicity was observed. Our preliminary results support the validity of this approach and the efficacy of this combination chemotherapy. A large-scale randomized study is warranted.