Data collected prospectively on over 3800 cadaveric renal transplants performed between June 1977 and July 1982 by the 41 member institutions of the South-Eastern Organ Procurement Foundation were analyzed to determine the influence of delayed graft function (DGF) on patient and graft outcome. Approximately 35% of first graft recipients and 47% of regrafted patients were found to have DGF, as determined by the necessity for dialysis at one week posttransplant. First-graft recipients with DGF tended to include more black recipients, patients with higher peak levels of panel reactive antibody (PRA), less use of antilymphocyte serum (ALS) posttransplant, slightly longer organ preservation times and the more frequent use of organs by ice alone. Multivariate (Cox) regression analysis considering DGF simultaneously with ten other potentially confounding variables showed a highly significant association between DGF and overall graft loss from all causes (P less than 10(-5], irreversible graft rejection (P less than 0.001) as well as patient death (P = 0.012). The differences in graft survival between first graft recipients with DGF (n = 961) versus those without DGF (n = 1769) at one and four years posttransplant were 46% +/- 2 vs. 60% +/- 1 and 28% +/- 3 vs. 40% +/- 2, respectively. The detrimental effect of DGF was highly significant irrespective of the source of donor organs or the type of preservation used. For first transplant recipients who recovered good graft function by one month following DGF (n = 564), there was a significant decrease in eventual graft survival, as compared with patients who had graft function at one month but no prior history of DGF (n = 1407; P = 0.008). However, patients with history of DGF who had good graft function at six months (n = 361) showed no significant difference in longer-term graft survival when compared with similar patients with good graft function at six months but no history of DGF (n = 912). Interestingly, first transplant recipients with DGF were found to have significantly better graft survival if they had received bilateral native nephrectomy at least one month prior to transplantation. These results indicate that delayed graft function following cadaver donor renal transplantation provides a significant risk for eventual graft and patients survival that is principally manifested during the first six months posttransplant. In addition, patients who recover graft function following DGF appear to also remain at higher risk for early graft loss, while pretransplant bilateral native nephrectomy may afford some protection against the detrimental effects of DGF.