Mitoxantrone (MTZ) has been shown to be effective in the treatment of newly diagnosed acute myeloblastic leukemia (AML). The objective of this randomized study was to evaluate the impact of mode of administration of MTZ on the response and recurrence rates in newly diagnosed patients with AML and to compare the toxicity patterns associated with bolus and continuous infusion (CI) of MTZ. From March 1987 to March 1994, 40 newly diagnosed patients with AML were randomized to receive either bolus or CI-MTZ, administered for 3 days at 10 mg/m(2)/day in combination with CI-cytarabine for 7 days at 100mg/m(2)/day. Patients achieving complete remission (CR) received two consolidation cycles followed by monthly maintenance cycles, aiming a total of 12 cycles of chemotherapy. Fifteen patients (75%) in the bolus arm and 16 patients (80%) in the CI arm achieved CR. There were no significant differences in rates of early death and time to myeloid recovery between the two groups. After 11 years from the initiation of the study, median disease-free survival (DFS) in bolus and CI groups were 19 and 29 months after a median follow-up of 10 and 14 months, respectively. DFS rates at 10 years were 16.7% in the bolus group and 28.6% in the CI group (p = 0.36). Overall survival (OS) rates during the same period were 10.7 and 21.3% in the bolus and CI groups, respectively (p = 0.26). No relapse was observed in either group after 4 years. In patients younger than 40 years of age, DFS and OS were found to be significantly longer in the CI arm (p = 0.02 and p = 0.03, respectively). Mild asymptomatic cardiotoxicity associated with a decrease of 10 to 20% in the ejection fraction occurred in a patient in CI-MTZ arm and in two patients in the bolus arm. None of these patients showed any evidence of cardiac failure during their subsequent follow-up. Grade III-IV alopecia (p = 0.05) and grade I-II hepatotoxicity (p = 0.01) were more frequent in the CI arm. A tendency for higher frequency of grade III-IV nausea was observed in the bolus arm (9 vs. 3%, p = 0.10). As a conclusion, bolus and CI administration of MTZ were equally effective and tolerated well. Development of new anti-leukemia agents with novel treatment approaches is still needed to improve the high relapse rates in patients with AML who do not have an HLA-matched donor. Copyright (C) 2004 John Wiley Sons, Ltd.