OBJECTIVE: To study the efficacy and toxicity of ganciclovir prophylaxis given at engraftment to cytomegalovirus (CMV)-seropositive, allogeneic bone marrow transplant recipients. DESIGN: A double-blind, placebo-controlled study. SETTING: The Fred Hutchinson Cancer Research Center, a referral marrow transplant center. PATIENTS: This study was conducted from November 1990 to August 1991. Ninety-three CMV-seropositive patients were entered into the study before marrow transplant, with 64 patients randomized to receive the study drug after marrow engraftment. Thirty-one patients received placebo, and 33 received ganciclovir. The dose was 5 mg/kg body weight administered intravenously twice daily for 5 days, followed by once daily until day 100 after transplant. MEASUREMENTS: Outcome variables measured were CMV infection, monitored by weekly cultures, and neutropenia, defined as an absolute neutrophil count of 0.750 x 10(-9)/L for 2 consecutive days. Cytomegalovirus disease and mortality were secondary end points. RESULTS: Fourteen (45%) placebo recipients developed CMV infection in the first 100 days after marrow transplant compared with one (3%) ganciclovir recipient (P < 0.001). Nine (29%) placebo recipients developed CMV disease compared with no cases in the ganciclovir group during the first 100 days (P < 0.001). Neutropenia occurred in 10 ganciclovir recipients (30%) compared with no cases in the placebo group during the period of observation (P = 0.001). In a separate analysis, patients on ganciclovir who became neutropenic were at greater risk (relative risk, 4.3; P = 0.02) for bacterial infection. Mortality between the two study groups did not differ statistically at 100 and 180 days. CONCLUSION: Ganciclovir given prophylactically after engraftment is effective in suppressing CMV infection and disease. Neutropenia is an important side effect of ganciclovir use and is associated with an increased risk for bacterial infection.