Of 2457 patients in the North American Pediatric Renal Transplant Cooperative Study registry who were followed for 5481 patient-years after the index transplantation, we observed 136 deaths, for an average annual rate of 24.8 deaths per 1000 patient-years. Death resulted primarily from infection (n = 55, 40%), cardiovascular causes (n = 28, 21%), hemorrhage (n = 16, 12%), and malignancies (n = 9, 7%). Cadaver-donor source was associated with greater mortality (6.7%) than a living-donor source (4.0%) (P < 0.005). Recipients aged 0-1, 2-5, 6-12, and 13-17 years old had mortality rates of 17.5, 8.0, 3.6, and 4.5%, respectively (P < .001). Mortality rates increased substantially when examined by recipient and cadaver donor ages (mortality rates of up to 45%), the greater the concordance between young donor and recipient ages. Interestingly, acute tubular necrosis and graft failure less than 30 days after transplantation (GH30) were each associated with markedly elevated mortality rates. (The risk ratio for ATN was 3.1 [P < 0.001] and for GF30 it was 6.4 [P < 0.001].) Mortality after transplantation was also affected by the underlying renal disease, with high mortality rates observed for oxalosis (n = 21, 33.3%), congenital nephrotic syndrome (n = 79, 15.2%), pyelo/interstitial nephritis (n = 54, 11.1%), and Drash syndrome (n = 14, 21.4%). When the joint effect of these risk factors was examined in a Cox proportional hazards model, young recipient age (0-1 years old) and GF30 were significant (P < .001) risk factors of mortality for recipients of living-donor organs. For recipients of cadaver kidneys, young recipient age--0-1 years old (P < .001) and 2-5 years old (P = .002)--ATN (P = .029), and GF30 (P < .001) were all significant risk factors. Recipient age is the major determinant of increased mortality after renal transplantation. Avoidance of acute tubular necrosis by reducing cold time and preventing early graft failure by better matching techniques in this vulnerable population may improve the mortality rate.