The antitumor activity of carboplatin (400 mg/m2 intravenously every 4 weeks) in advanced breast cancer was evaluated in two consecutive trials enrolling patients with and without prior exposure to chemotherapy, respectively. All patients had measurable disease in at least one site. The first trial included patients who had received prior chemotherapy (adjuvant, neoadjuvant, or chemotherapy for metastatic disease). All but one of these patients had previously received doxorubicin-containing combinations. There were no objective responses among the first 14 evaluable patients, although 7 of them had stable disease, including 2 with minor responses. The second trial, carried out with patients who had no prior exposure to chemotherapy, is ongoing. Currently, 6 of 19 evaluable patients have obtained a complete (1) or partial (5) response to carboplatin, resulting in an overall response rate of 32%. In both studies, toxicity was mild, mainly consisting of emesis (88%), leukopenia (22%), and thrombocytopenia (12%). Thus, by standard criteria, carboplatin was not found to be active in breast cancer patients with prior exposure to chemotherapy. Preliminary results in patients without such exposure are encouraging, although additional patients are needed to confirm these data.