Oral valacyclovir for cytomegalovirus (CMV) prophylaxis in bone marrow transplantation (BMT) was investigated in a randomized, double-blind, acyclovir-controlled, multicenter clinical trial in recipients of allogeneic BMT who were CMV seropositive (or donor positive) before transplantation and were aged 13 years or older. Patients were randomized before BMT. All initially received intravenous acyclovir (500 mg/m(2)) 3 times daily until day 28 after transplantation or after discharge, then oral valacyclovir (2 g) or acyclovir (800 mg) 4 times daily until week 18 after transplantation. Evidence of CMV infection, CMV disease, and death were documented for 22 weeks. Primary end points were time to CMV infection (detection of CMV in blood, broncho-alveolar lavage) or CMV disease and survival. Preemptive CMV therapy was permitted. Seven hundred twenty-seven patients were evaluable for efficacy. After the administration of intravenous acyclovir, valacyclovir was significantly more effective than oral acyclovir in reducing the incidence of CMV infection. CMV infection or disease developed in 102 (28%) valacyclovir patients, compared with 143 (40%) acyclovir patients (HR, 0.59; 95% CI, 0.46-0.76; P <.0001). Survival did not differ between treatments (76% and 75% in the valacyclovir and acyclovir groups, respectively). The safety of oral valacyclovir was similar to that of high-dose oral acyclovir. Valacyclovir was more effective than acyclovir in preventing CMV reactivation in BMT recipients and showed a similar safety profile. CMV disease incidence was low, and no differences were observed between oral valacyclovir and acyclovir. Survival was similar in each group. Valacyclovir prophylaxis provides a clinically valuable intervention but must be part of an overall strategy for CMV prevention in BMT.