Background. Kidney transplants using older donors are becoming increasingly accepted as a strategy for alleviating the growing donor organ shortage. Most studies to date have shown decreased graft survival associated with the use of older cadaver donors; however, studies on the effect of living donor age on graft survival are less clear-cut. Methods. We studied the effect of donor age on patient and graft survival after 1126 consecutive cyclosporine-treated primary kidney transplants performed between January 1, 1985 and December 31, 1995. Of these grafts, 598 were from living donors (74 from donors >55 years old) and 528 from cadaver donors (54 from donors >55 years). We calculated actuarial patient survival, graft survival, and death- censored graft survival for recipients of both living donor and cadaver kidneys. Living donors were then further divided by HLA mismatch (0 vs. 1 - 6) and the presence or absence of an acute rejection episode. Multivariate analysis of factors associated with decreased graft survival was performed for recipients of both living and cadaver donor kidneys. Factors included for analysis were donor age >55 years, recipient age >50 years, the presence of diabetes mellitus, HLA mismatch (0 vs. 1 - 6), and the presence of an acute rejection episode. Results. For cadaver kidneys, univariate analysis indicates that both overall (P=0.004) and death-censored (P=0.001) graft survival was significantly better with younger cadaver kidneys. This is supported by our multivariate analysis, which shows that cadaver donor age >55 years is an independent predictor of poor actuarial graft survival (P=0.0003). For living donor kidneys, univariate analysis also indicates that both overall (P=0.045) and death-censored (P=0.005) graft survival was significantly better with younger living donor kidneys. However, in the absence of acute rejection, 10-year death-censored graft survival for patients with older vs. younger living donor kidneys was 93% vs. 94%, whereas in the presence of one or more acute rejection episodes, 10-year death- censored graft survival dropped markedly to 39% with older and 54% with younger living donors. Kidneys from living donors >55 years had significantly better long-term graft survival than cadaver donors >55 years (P=0.012) and had comparable graft survival to younger cadaver donors. In contrast to our univariate analysis, multivariate analysis of our living donor data shows that decreased actuarial living donor death-censored graft survival was significantly associated only with the presence of one or more acute rejection episodes (P<0.0001). Living donor age >55 years was not independently associated with decreased graft survival. Conclusions. Ours is the largest single-center study of outcome for recipients of kidneys from living donors >55 years. Using univariate analysis, we have shown that graft survival of kidneys from older living donors is significantly better than that of kidneys from older cadaver donors and is comparable to that of kidneys from younger cadaver donors. Using multivariate analysis, we have shown that the presence of one or more acute rejection episodes significantly shortens both cadaver and living donor long-term graft survival. Most significantly, we have shown that, although the use of kidneys from cadaver donors >55 years is associated with significantly decreased long-term graft survival, no such association exists for recipients of kidneys from living donors >55 years. We feel that our data support the continued use of kidneys from older living donors.