Background: HSCT has been reported to be a treatment option for elderly patients with AML. Here we analyze the outcome of CT compared to HSCT in elderly AML patients according to genetic risk groups defined by the European Leukemia Net. Methods: By May 2015, 492 of 789 eligible patients (62%) from the AML 2004 study group entered CR and were assigned to CT (n = 205), matched (n = 119) or one antigen/allele mismatched HSCT (n = 31) depending on donor availability. Median age was higher (68 vs 66 years; p < 0.0005) and intermediate risk (IR) II and high risk (HR) cytogenetics were less frequent (p = 0.002) in CT than in HSCT. The interval CR - CT was significantly shorter than CR - HSCT (43 vs 65 days; p < 0.0005]. AML type, NPM1 and FLT3 status were comparable. Results: Patients receiving matched HSCT had longer LFS than those receiving CT (25±5% vs. 14±3% at 9 years; p < 0.001). As expected, RI was lower (42±5% vs 78±3%; p < 0.0001) and NRM was higher (34±5% vs 8±2%; p < 0.0001) at 9 years in HSCT than in CT patients. LFS differed between HSCT and CT for IR I (28±8% vs. 16±4% at 9 years; p = .007), IR II (30±10% vs. 4±4% at 7 years; p = .08) and HR patients (12±7% vs. 0±0% at 7 years; p < .05). RI was decreased in HSCT vs. CT for IR I (35±7% vs. 74±4% at 9 years; p < .0001), IR II (38±10% vs. 96±5% at 7 years; p < .0001) and HR (59±11% vs. 96±7% at 9 years; p < .004). These effects outweigh the higher NRM in HSCT for IR I (37±9% vs. 10±3% at 9 years; p = .0005), IR II (33±10% vs. 0% at 7 years; p = .003) and HR (29±10% vs. 4±4% at 9 years; p = .11). HSCT was an independent prognostic factor for NRM (p < 0.005), LFS and RI (both p < 0.0005). Genetic risk group (p < 0.01) was the only factor associated with OS. Conclusions: HCT improves LFS in AML CR1 patients between 60 and 75 years of age in genetic risk groups from IR I to HR. Clinical trial information: NCT01497002