Cyclosporine-treated recipients of primary cadaver donor renal transplants had a one-year graft survival rate of 79% if they received pretransplant blood transfusions (n = 5308). The one-year survival rate for nontransfused recipients (n = 709) was significantly lower at 69% (P less than 0.001). The transfusion effect was larger in black recipients (a 17% difference) than in white recipients (5%). The effect was also larger in recipients of grafts poorly matched for HLA-A, B, -B, DR, or -DR antigens than in recipients of well-matched grafts. Transfusions did not significantly improve graft survival in recipients with zero or one HLA-A, B or -B, DR, or zero -DR-mismatched grafts. However, transfusions accounted for increases of 10%, 14%, and 17% in patients receiving grafts mismatched at 2, 3, or 4 HLA-B, DR antigens, respectively. Several factors including cyclosporine and HLA matching have contributed to improving graft survival rates in nontransfused recipients. Sensitization was noted in 20% of transfused patients awaiting primary renal transplants in Southern California, as compared with 10% in transplanted patients, suggesting a tendency to transplant nonsensitized patients. Of the sensitized patients, 75% were female. Based on these data, we suggest that high survival of primary kidney allografts in the cyclosporine era can best be maintained by the continued use of pretransplant transfusions for the majority of recipients--or, alternatively, by HLA matching for patients who are at higher risk of becoming sensitized.