1. Second-graft survival rates were 67% at 1 year and declined to 40% at 5 years, consistently 10% lower than those of first grafts. First- and third-graft survival rates were 77% and 58% at 1 year, respectively. The difference was apparent at 3 months. 2. Second transplants with good 1-month function had an 82% 1-year graft survival rate compared to 86% for first transplants. With good 3-month function, they had 89% 1-year survival, comparable to 91% for first transplants. 3. First-graft duration had a high prognostic value for second-graft survival. Second transplants with first-graft duration less than 3 months had 57% 1-year graft survival whereas those with more than 12 months had 75% 1-year survival. The difference was significant up to 5 years. Half-life after 1 year was 5.9 for acute responders and 5.3 for chronic responders. 4. The FCXM was a useful screening tool for second transplants. Second transplants with positive FCXM had 61% 3-month survival and those with a negative crossmatch had 82%. A positive FCXM with short first-graft duration had 48% 3-month survival and should be a contraindication for retransplantation. 5. Responder status was associated with HLA-mismatches in the first transplants. The 0-HLA-mismatched first grafts resulted in acute responders for 45-58%, which was significantly lower than 62-69% of totally HLA-mismatched first grafts. Totally HLA-mismatched first grafts were related to 3-18% lower survival for acute responders. 6. Acute responders benefited more from HLA matching than chronic responders. In acute responders, 0-HLA-A,B-mismatched grafts had a 72% 1-year graft survival rate, and 4 mis-matched had 58%. At 3 years, they had 59% and 42%, respectively. 7. HLA-DR1 recipient phenotype was associated with a 10-15% increased survival rate. DR1-positive recipients had 73% 1-year survival compared to 63% of DR1-negative recipients. 8. Never-transfused second-transplant recipients had a 59% 1-year survival rate compared to 69% for transfused. Transfusions of 1-4 units were sufficient to obtain a transfusion effect. 9. The antihuman immunoglobulin crossmatch test was associated with a 3-8% better 2-year graft survival in first and second transplants. There was no short-term beneficial effect. This assay might have detected false-negative crossmatch cases not seen by NIH or one-wash.