The preliminary results of a controlled study with two independent combinations in metastatic breast cancer are reported. The first combination (Therapy A: 41 patients) consisted of Adriamycin (ADM) and vincristine (VCR), while the second combination (Therapy B: 41 patients) included cyclophosphamide (CTX), methotrexate (MTX), and 5-fluorouracil (FU), designated "CMF." Both treatments were administered as intermittent cycles in patients previously untreated with chemotherapy. After eight cycles, responsive patients on Therapy A were crossed over to Therapy B to avoid the risk of cardiomyopathy. In both groups, crossover was carried out at the time of progression or relapse. In the group given ADM plus VCR, the dominant site of disease was in soft tissues in 56%, in viscera in 22%, and in bones in 22%. These findings were present in 51%, 24%, and 24%, respectively in the group started on CMF. The comparison of response after primary treatment in patients receiving a minimum of two cycles failed to show a significant difference between Therapy A and Therapy B (overall response 58% vs. 65%). When the response rate was calculated only in patients who had completed the first eight cycles of therapy, these findings were 87% and 93%, respectively. Three patients receiving treatment A (8%) AND 4 patients given Treatment B (10%) achieved complete remission. The highest incidence of response was observed for soft tissue lesions (70% vs. 76%). However, complete or partial bone recalcification was seen in 33% and 24%, respectively. The duration of response was found to be longer in patients who crossed over to CMF after eight cycles of ADM plus VCR, as compared to those started and continued on CMF. No cross-resistance was observed after crossover for progressive disease. Both incidence and degree of side effects were found acceptable, and no drug-related death was seen. Virtually all patients were treated on an outpatient basis.