Preemptive antiviral therapy is often employed for CMV prevention following allogeneic BMT. Two common strategies are a screening bronchoscopy for CMV post-BMT or regular CMV antigenemia testing with ganciclovir administration for a positive result. In a randomised trial, we prospectively compared the efficacy of these two preemptive strategies. Consecutive patients were randomised to either a bronchoscopy for CMV on day 35 post BMT or weekly CMV antigenemia testing. If the bronchoscopy was positive for CMV, patients received preemptive ganciclovir for 8-10 weeks. If the antigenemia was positive for CMV, patients received a minimum of 2 weeks of preemptive ganciclovir. The primary endpoint was the development of active CMV disease. One hundred and eighteen allogeneic BMT patients were enrolled (60 in the antigenemia arm and 58 in the bronchoscopy arm). The two groups were comparable with respect to baseline demographic data, underlying disease, conditioning regimen, and immunosuppression. Active CMV disease developed in 7/58 (12.1%) patients in the bronchoscopy arm vs 1/60 patients (1.7%) in the CMV antigenemia arm (P = 0.022). Based on the screening test, 13.8% of patients received preemptive ganciclovir in the bronchoscopy arm vs 48.3% of patients in the antigenemia arm (P < 0.001). There was no significant difference in the rate of graft-versus-host disease, bacteremia, invasive fungal infections or mortality between the two groups. Preemptive therapy based on regular CMV antigenemia monitoring is superior to screening bronchoscopy for the prevention of CMV disease after allogeneic BMT.