The treatment of older patients with acute myeloid leukemia (AML) remains unsatisfactory, with complete remission (CR) achieved in only approximately 50% and long-term disease-free survival in 10% to 20%. Three hundred eighty-eight patients (60 years of age and older) with newly diagnosed de novo AML were randomly assigned to receive placebo (P) or granulocyte-macrophage colony-stimulating factor (GM-CSF) or GM in a double-blind manner, beginning 1 day after the completion of 3 days of daunorubicin and 7 days of cytarabine therapy. No differences were found in the rates of leukemic regrowth, CR, or infectious complications in either arm. Of 205 patients who achieved CR, 169 were medically well and were randomized to receive cytarabine alone or a combination of cytarabine and mitoxantrone. With a median follow-up of 7.7 years, the median disease-free survival times were 11 months and 10 months for those randomized to cytarabine or cytarabine/mitoxantrone, respectively. Rates of relapse, excluding deaths in CR, were 77% for cytarabine and 82% for cytarabine/mitoxantrone. Induction randomization had no effect on leukemic relapse rate or remission duration in either postremission arm. Because cytarabine/mitoxantrone was more toxic and no more effective than cytarabine, it was concluded that this higher-dose therapy had no benefit in the postremission management of older patients with de novo AML. These results suggest the need to develop novel therapeutic strategies for these patients. (Blood. 2001;98:548-553)