1. Long-term graft survival is characterized by failure rates that are essentially constant after about a year or 2. The rate has not varied appreciably among cohorts defined in terms of transplant year. 2. Transplants with an excellent clinical course throughout the first 3 months (about 2-thirds of cadaver-donor first transplants) have substantially better long-term graft survival than those with a less favorable early course. 3. Having an excellent early clinical course did not remove the effects of the several factors related to long-term survival. 4. Use of cyclosporine did not relate to long-term survival. 5. Factors related to short-term graft survival (1 year), generally related to long-term survival and in the same direction. Cyclosporine use was a major exception. 6. The dominant factor for long-term survival was tissue matching as reflected in donor categories: HLA-identical sibling, parent, and cadaver. 7. Black recipients had a decidedly poorer long-term survival than recipients of other races/ethnicities. 8. Transplant center was a dominant factor in long-term graft survival. 9. Multivariable analysis did not materially alter the findings obtained from factors considered separately. This was attributable to the sensitivity associated with small standard errors resulting from the large number of cadaver-donor transplants and the combination of smaller numbers and longer survival of living-related transplants.