The effects of cytomegalovirus (CMV) infection on patient and allograft survival were determined in 1245 renal transplant recipients from 46 transplant centers. When an antilymphocyte preparation was administered to cadaveric allograft recipients, those at risk for primary CMV had a worse outcome than similar patients treated with prednisone and azathioprine (53.1% alive at 6 months with a functioning allograft vs. 70.8%, P = .05) or patients at risk for reactivation CMV (53.1% vs. 71.1%, P = .035). Patients at risk for reactivation CMV had a better outcome if they received an antilymphocyte preparation (71.1% vs. 60.8%, P less than .01). The type of immunosuppression had no effect on patients without CMV. Living-related donor transplantation was not significantly influenced by CMV or type of immunosuppression. We conclude that CMV infection is strongly influenced by the form of immunosuppression employed, and that both are important determinants of the outcome of cadaveric renal transplantation.