Predictive factors of response to hypomethylating agents (HMA) in elderly acute myeloid leukemia (AML) patients remain unclear in the real-life setting and no direct comparison between azacitidine (AZA) and decitabine (DEC) has been carried out. We retrospectively evaluated 110 AML patients treated with HMA (78 AZA, 32 DEC) as first-line therapy outside of clinical trials. Median age was 75 years (range 58-87). The median overall survival (OS) of the entire cohort was 8.0 months (95% CI 6.1-10), without significant differences among the subgroups: AZA 8.8 months vs DEC 6.3 months (p = 0.291). HMA treatment yielded an overall response rate (ORR) of 40% (AZA 37% vs DEC 47%, p = 0.237). A stable disease (SD) after 4 HMA cycles was not associated with a worse survival outcome compared with an early optimal response. Factors independently associated with a better OS were transfusion independence during treatment (p = 0.049), achievement of an optimal response to treatment (p < 0.001), and a baseline hemoglobin level ≥ 9.25 (p = 0.018). A bone marrow (BM) blast count ≥ 30% (p < 0.001) and a therapy-related AML (p = 0.008) remain poor survival predictors. Of the available biologic features, an adverse risk category according to the ELN classification was significantly associated with a shorter survival over the intermediate risk category (p = 0.034). Disease progression remains the primary cause of death. Infectious complications were more severe (p = 0.036) and occurred earlier (p = 0.006) in the DEC group compared with that of the AZA group. In conclusion, clinical prognostic factors associated to response and survival have been identified without significant associations concerning overall outcomes between the two HMAs.