Sixty-eight patients with previously untreated metastatic breast cancer were randomly assigned to receive either 'classical CMF' (orally administered cyclophosphamide) or new intravenous administration of all drugs every 3 weeks. Overall response rates of 44.5% (95% CI: 28-62%) and 39% (95% CI:24-54%) were observed with classical and new CMF, respectively. The time to progression and overall survival were also similar: hematologic toxicity was mild in both groups, but the tolerance and patient compliance was generally better for the new CMF. The patients treated with classical CMF received a significantly higher drug intensity; dose intensity, however, had no impact on clinical results. In conclusion, our data suggest that the new CMF schedule has comparable activity to classical CMF but better patient compliance. A brief review of recent clinical studies regarding dose-effect relationships in breast cancer is included.