A variety of refinements in the care of kidney transplant recipients have been instituted over the past decade. The authors studied the overall impact of these refinements on kidney allograft losses at a single institution. To do this they compared the causes and rates of graft loss for primary kidney transplants in the 1970s (January 1, 1970 to December 31, 1979;n = 1012; 657 nondiabetics, 355 diabetics; 617 living donors, 395 cadaver donors) versus the 1980s (January 1, 1980 to December 31, 1989; n = 1,384; 756 nondiabetics, 628 diabetics; 740 living donors, 644 cadaver donors). Overall patient survival improved significantly, with rates at 1, 5, and 10 years of 94%, 84%, and 68% for the 1980s, compared with 86%, 69%, and 57% for the 1970s (p < 0.001). Actuarial graft survival also improved significantly, with rates at 1, 5, and 10 years of 86%, 71%, and 52% for the 1980s, compared with 73%, 58%, and 43% for the 1970s (p < 0.001). This improvement occurred even though there were proportionately more cadaver donors and diabetic recipients in the 1980s. For both decades combined, 24% of the lost grafts were due to chronic rejection, 18% to cardiovascular causes of death with function, 13% to infectious causes of death with func-tion, and 11% to acute rejection. The overall gain in graft survival rates in the 1980s was principally due to fewer cases of acute rejection and fewer infectious deaths. Improvement in graft survival due to the two leading causes - chronic rejection and cardiovascular causes of death - was relatively small, if any. These data indicate that future kidney transplantation research should emphasize prevention of chronic rejection and cardiovascular death.