In a phase II study, 32 patients with advanced breast cancer previously unexposed to palliative cytotoxic chemotherapy were treated with amonafide, 800-900 mg intravenously over 3 h repeated every 4 weeks. Objective response was seen in 8 patients including 1 complete response, 10 patients had stable disease and 14 patients progressed so the overall response was 25% (95% confidence interval, 11-43%). The most frequently encountered side-effects were haematological (granulocytopenia > or = WHO grade 3 was encountered in 7/24 patients at 800 mg/m2 and in 3/8 patients at 900 mg/m2 amonafide) and nausea/vomiting (62%), despite prophylactic use of ondansetron. Non-haematological severe adverse reactions included neurotoxicity WHO grade 3 in 1 patient and orthostatic hypotension WHO grade 4 in another. In summary, the results of this trial suggest a limited therapeutic index of amonafide if used at this dose with this administration schedule.