Background: Controlling cytomegalovirus (CMV) infection through prophylaxis or pre-emptive therapy remains an important contributor to outcomes after allogeneic hematopoetic stem cell transplant (alloHCT). Predicting clinically significant CMV infection (csCMVi) after day 100 remains a challenge. Methods: We examined the abilty of the QuantiFERON-CMV assay (QFN-CMV) at day 100 (d100) and day 150 (d150) after alloHCT to predict csCMVi after these time points, with median follow-up of 3.1 years (range 1.3–4.3 years). Results: In 46 transplants (donor seropositive (D+) recipient seronegative (R-) = 12, D+R+ = 25, D-R+ = 9; matched related = 13, unrelated donor = 32, haploidentical = 1), for the prediction of freedom from csCMVi >d100, QFN-CMVd100 (positive compared to negative/indeterminate) had sensitivity 62% (23/37), specificity 100% (9/9), positive predictive value 100% (23/23), and negative predictive value 39% (9/23). For the prediction of freedom from csCMVi >d150, QFN-CMVd150 (positive compared to negative/indeterminate) had sensitivity 62% (18/29), specificity 83% (5/6), positive predictive value 95% (18/19), and negative predictive value 31% (5/16). Conclusion: Positive QFN-CMV at d100 and d150 strongly predicted freedom from csCMVi after these time points. QFN-CMV could be utilized to predict the need for pre-emptive therapy and CMV viral load monitoring after day 100 post-alloHCT.