Background. Live donor renal transplantation (LRT) now comprises more than 40% of all kidney transplants performed in the United States. Many patients on the cadaveric waiting list have a prospective live kidney donor. This study determines whether cadaveric donor renal transplantation (CRT) can demonstrate better outcomes than LRT. Methods. From the United States Renal Data System registry, 31,909 adult recipients of a first-time kidney transplant from 1995 to 1998 were analyzed. Recipients were followed until December 31, 2000. Results. CRT, more human leukocyte antigen (HLA) mismatches, increased donor age, cold ischemia time greater than 24 hr, African American recipient, and a history of diabetic nephropathy all increased the risk of graft failure, return to dialysis, and death. Nevertheless, in specific circumstances, CRT could provide better outcomes than LRT. For example, in recipients aged 18 to 59 years with a hypothetical live kidney donor aged 50 years and four HLA mismatches, the relative risk of graft loss with LRT is comparable or increased compared with CRT if the cadaveric kidney donor is much younger or with fewer HLA mismatches. On the other hand, for recipients aged 60 years or older, CRT never provides better outcomes than LRT. All analyses were adjusted for recipient race, gender, and history of diabetic nephropathy. There were no significant interactions among donor type, HLA mismatches, donor age, and cold ischemia time. Conclusions. The elderly recipient with an imminent LRT should never be offered CRT. A combination of recipient and donor factors can make CRT preferable to LRT in younger patients.