Letermovir is the first approved drug for cytomegalovirus (CMV) infection prophylaxis in adult CMV-positive patients undergoing allogeneic hematopoietic cell transplantation (allo-HCT). Since CMV infection risk varies from patient to patient, we evaluated whether a risk-based strategy could be effective. All consecutive CMV-positive adult patients (median age: 56.0 years, IQR: 43.5-64.0) who underwent allo-HCT between 2015 and 2021 (n=316) were included. Letermovir was not used in Period 1 (2015-2017, n=186), whereas during Period 2 (2018-2021, n=130), letermovir was used in high-risk patients but not in low-risk patients, except in those receiving corticosteroids. In high-risk patients, the incidence of clinically significant CMV infection (csCMVi) in Period 2 was lower as compared to Period 1 (p<0.001) by week 14: 10.5% (95% CI 4.6-19.2) vs 51.6% (41.0-61.3) and week 24: 16.9% (8.9-27.0) vs 52.7% (42.0-62.3), respectively. In low-risk patients, although only 28.6% of patients received letermovir in Period 2, csCMVi incidence was also significantly lower (p=0.003) by week 14: 7.9% (2.9-16.3) vs 29.0% (20.2-38.5) and week 24: 11.2% (4.9-20.5) vs 33.3% (23.9-43.0). Among low-risk patients who did not receive letermovir (n=45), 23 (51.1%) patients experienced transient positive CMV DNA without csCMVi by week 14, while it remained negative in 17 (37.8%) patients. In both risk groups, the two periods were comparable for CMV disease, overall survival, progression-free survival, relapse, and non-relapse mortality. We concluded that a risk-based strategy for letermovir use is an effective strategy which maintains the high efficacy of letermovir in high-risk patients but allows some low-risk patients not to use letermovir.