One hundred twenty-two patients with advanced adenocarcinoma of the breast were randomized to receive adriamycin (AD) alone or a combination of VP-16 plus lower dose adriamycin (VAD). The patients were stratified to good and poor risk. The starting dose (day 1) of AD was 60 mg/m2 for good risk and 45 mg/m2 for poor risk. The starting dose of the VAD combination for the good-risk patient was VP-16, 75 mg/m2 daily x 5 plus adriamycin 35 mg/m2. The poor-risk dose for VAD was VP-16, 50 mg/m2 daily x 5 plus adriamycin, 30 mg/m2 on day 1. The total dose of AD was 450 mg/m2 on both arms. The patients who were on the VAD arm continued on VP-16 maintenance. Both arms were repeated every 21 days. There were 54 evaluable patients on the adriamycin arm and 52 evaluable patients on the VAD arm. Both arms were similar with regard to age, menopausal status, performance status, and prior hormonal therapy. More hematologic toxicity was seen in the adriamycin arm. Complete responses were observed on both arms, three (5%) with adriamycin and three (5%) with combination. Eleven (19%) and ten (18%) partial responses were observed with the adriamycin and VP-16 plus adriamycin, respectively. AD produced more stable disease than VAD (41% vs. 29%). Complete responses were seen only in the good-risk patients. Time to progression was delayed on the combination arm (P = 0.02). The survival in both arms was similar. The addition of VP-16 to adriamycin does not offer an important clinical advantage.