SUMMARY Two factors which have gained attention as possible contributors to success of renal allografts are freedom from infection with cytomegalovirus (CMV) after transplant and administration of multiple blood transfusions pretransplant. In order to determine the interrelationship of these two variables, we analyzed 55 recipients of well matched (at least two antigens) cadaveric kidneys. In this study, absence of CMV infection and receipt of multiple transfusions both provided favorable outcomes. When patients were grouped by both factors, those who were free of infection and received multiple transfusions did significantly better than any other combination. Infection with CMV decreased the frequency of allograft survival in multiply transfused patients to a point intermediate between the above group and those who had not been multiply transfused. Since CMV infection can be predicted by measurements made pretransplant, and since up to one-quarter of CMV infection which develops post-transplant is transmitted with the allograft, administration of multiple transfusion to all patients and the use of donors who are free of latent CMV infection for CMV antibody-negative potential recipients should increase allograft success. For those potential recipients who already have latent CMV infection, further study will be necessary to determine what stimuli can be given pretransplant to prevent the interference with the beneficial effect of multiple transfusions Recent reports have emphasized factors other than HLA-A and B locus antigen match which influence the survival of renal allografts. Two factors that have received much attention are viral infection and multiple pretransplant transfusions Opelz et al. (1) in 1973 demonstrated that allograft survival was significantly better in patients who had received multiple blood transfusions before transplant. Although there was concern that multiple transfusions would produce cytotoxic antibody to HLA which in turn would lead to hyperacute or accelerated rejection, it seems clear now that the routine performance of a pretransplant crossmatch between donor and recipient avoided this complication. The protection afforded by multiple transfusions has subsequently been confirmed in a number of studies (2-6). Frozen cells have been shown to be as protective as fresh blood (7) It was recognized in 1964 that infection with CMV was common in renal transplant patients (8, 9). Subsequently, it has been demonstrated that CMV shedding occurs in 40 to 80% of transplant recipients (10-12) and that it is more common in patients who have antibody to CMV prior to transplantation than in those who lack antibody (10-13). The explanation for this became apparent when evidence was presented that CMV infection could be transmitted with a renal allograft (70,12). In fact, it has become clear that all post-transplant virus shedding can be predicted provided that donors as well as the recipients are tested by the fluorescent antibody technique (12) In subsequent studies, Betts et al. (14) and Suwansirikul et al. (15) showed that if CMV infection were acquired as a primary infection from the allograft then the CMV infection syndrome is common, whereas if the patient had antibody prior to transplant and shed virus after transplantation (reactivation infection), then the CMV infection syndrome was uncommon In 1970 Simmons et al. (16) suggested a relationship between viral infection and allograft rejection. In 1974, Lopez et al. (17) emphasized the importance of CMV. However, there is a difference of opinion about the role of CMV. Some investigators have found no relationship (10, 15). Others concluded that the activation of infection was the result of the rejection process (18). Still others noted a clear association between CMV infection and initial rejection episodes but found no long-term effect (19). Lopez et al. (17) noted a clear association but were unable to determine whether the infection was cause or effect. On the other hand, Simmons et al. (11) reported data which suggested a causal role for CMV infection. Betts et al. (14) pointed out that primary infection occurring in children receiving parental kidneys led to excess allograft loss and May et al. (20) found that CMV infection had more influence on allograft loss than did HLA match In that earlier study of 85 patients at the University of Rochester, it was found that infection with CMV had a more pronounced influence on allograft survival than did HLA antigen match. Furthermore, our data showed that CMV infection was not a consequence of rejection but could be predicted by measurements made before transplantation. In extending the study we have examined the interrelationship of the effect of multiple transfusions and of CMV infection on allograft success. In this report we have evaluated the combined effect of these variables on allograft outcome. This presentation is an analysis of these factors