The optimal intensity of conditioning for allogeneic hematopoietic stem cell transplantation (HCT) in acute myeloid leukemia (AML) remains undefined. Traditionally, myeloablative conditioning regimens improve disease control, but at the risk of greater nonrelapse mortality. Because fludarabine with myeloablative doses of intravenous busulfan using pharmacokinetic monitoring has excellent tolerability, we reasoned that this regimen would limit relapse without substantially elevating toxicity when compared to reduced intensity conditioning. We retrospectively analyzed 148 consecutive AML patients in remission receiving T cell replete HCT conditioned with fludarabine and intravenous busulfan at doses defined as reduced (6.4 mg/kg; FluBu2, n = 63) or myeloablative (12.8 mg/kg; FluBu4, n = 85). Early and late nonrelapse mortality (NRM) was similar among FluBu4 and FluBu2 recipients, respectively (day + 100: 4 vs 0 %; 5 years: 19 vs 22 %; p = 0.54). NRM did not differ between FluBu4 and FluBu2 in patients >50 years of age (24 vs 22 %, p = 0.75). Relapse was lower in recipients of FluBu4 (5 years: 30 vs 49 %; p = 0.04), especially in patients with poor risk cytogenetics (22 vs 59 %; p = 0.02) and those >50 years of age (28 vs 51 %; p = 0.02). Overall survival favored FluBu4 recipients at 5 years (53 vs 34 %, p = 0.02), a finding confirmed in multivariate analysis (HR: 0.57; 95 % CI: 0.34-0.95; p = 0.03). These data suggest that myeloablative FluBu4 may provide equivalent NRM, reduced relapse, and improved survival compared to FluBu2, emphasizing the importance of busulfan dose in conditioning for AML.