Taxol, an antimicrotubule agent, has shown promise for efficacy in treatment of breast cancer, but severe hypersensitivity reactions led to cessation of many phase I clinical trials. Consequently, investigators and the National Cancer Institute recommended that phase I and II studies of this agent use 24-hour infusions and antiallergic medications. Using a premedication regimen effective in preventing hypersensitivity reactions, we have performed a phase II trial of taxol in patients with metastatic breast cancer. Taxol was administered to 25 patients at a dose of 250 mg/m2 by 24-hour infusion every 21 days. These patients had received only one prior chemotherapy regimen, either adjuvant to surgery or for metastatic disease; all but two had received doxorubicin. In 60% of the patients, the dominant site of disease was the viscera. All patients were assessable. In April 1991, at a median time on study of 9 months (range, 5-13+ months), the objective response rate was 56% (12% complete and 44% partial; 95% confidence interval, 35%-76%). Disease progressed in only 8% of the patients. The median number of courses of therapy was 11. Granulocytopenia was the dose-limiting toxic effect, but neutropenia with fever occurred in only 5% of 232 courses. A chronic glove-and-stocking neuropathy developed in most patients, but no allergic reactions occurred. We conclude that taxol is an active agent in the treatment of metastatic breast cancer and that it warrants continued study. Currently, we are conducting a phase I trial of taxol plus doxorubicin. Future trials should address the optimal effective dose, the optimal sequencing of combinations, mechanisms of drug resistance in tumors, and dose-limiting toxic effects (particularly cardiac toxic effects of taxol given as a single agent or in drug combinations).