The activity of single-agent paclitaxel (Taxol; Bristol-Myers Squibb Company, Princeton, NJ) has been documented in untreated and previously treated metastatic breast cancer, including both patients with anthracycline-resistant disease and those with extensive pretreatment. Such activity has prompted investigations of the optimal doses and schedules of paclitaxel/doxorubicin combinations. With one exception, paclitaxel has been administered as either a 24- or a 3-hour infusion, while the administration times for doxorubicin vary from bolus injection to 72-hour infusion. Results of these completed phase I and II trials are reviewed. Also reported are two European trials that have achieved promising results. In Milan, a phase I/II trial has shown a preliminary response rate exceeding 90% in 32 chemotherapy-naive patients treated with an alternating schedule of paclitaxel given over 3 hours and intravenous bolus doxorubicin. At doses of paclitaxel 200 mg/m2 and doxorubicin 60 mg/m2, the dose-limiting toxicities were neutropenia, oral mucositis, myalgias, and peripheral neuropathy. Congestive heart failure occurred in six patients. A phase I/II study of a 30-minute doxorubicin infusion preceding a 3-hour paclitaxel infusion every 3 weeks in minimally pretreated patients also is reported. Of 29 patients evaluable for response, 17 have achieved partial responses and seven complete responses, for an overall response rate of 83% (95% confidence interval, 79% to 99%). Toxicities observed were grades 3 to 4 neutropenia and moderate paresthesias, nausea/vomiting, alopecia, myalgia, and mucositis. Cardiotoxicity also occurred, as 15 patients had a significant decrease in left ventricular ejection fraction measured by isotope cardiography. Six of these developed congestive heart failure. This effect has been observed only in studies using short infusions of both drugs, and it is now being investigated whether lowering the peak doxorubicin concentration will preclude it.