Actuarial survival analysis of 889 cadaver transplant patients between 1972 and 1981 in Michigan reveals functional graft (P = .0003) and patient (P = .004) survivals are improved when donors and recipients are of the same race (black or white). The Cox regression model for multifactorial analysis confirms the significantly lower graft survival for the mixed racial combination group with a relative risk of 1.27 (P less than .05). By this analysis, other significant factors adversely affecting the graft survival rates include diabetes as a cause of end-stage renal disease, earlier date of transplantation, shorter duration of prior dialysis, and a significant center effect. Patient survival has a significantly greater relative risk for the black to white racial combination, diabetes, earlier calendar year of transplantation, and age of patient. While the mixed racial group was slightly older (delta = 2 years), had more hypertension, less glomerulonephritis, and more HLA mismatches, our analysis by the Cox regression model suggests that these factors played only minor roles (P greater than .05) regarding graft survival rates. Therefore, our data suggest that independent of several other factors, cadaver kidneys have a better functional outcome when they are transplanted into recipients of the same race.