The Collaborative Corneal Transplantation Studies (CCTS) were designed to evaluate the effect of donor-recipient histocompatibility matching and crossmatching on the survival of corneal transplants in high-risk patients. Corneas were allocated to the 419 patients in the double-masked Antigen Matching Study on the basis of serologically defined HLA-A, -B and HLA-DR antigen match. ABO blood group compatibility was determined but not used for recipient selection. The 37 patients in the Crossmatch Study were randomly assigned to receive a cornea from either a positively or negatively crossmatched donor. All patients received topical steroid therapy according to a standard protocol. Matching for HLA-A, -B and HLA-DR antigens had no effect on overall graft survival, the incidence of irreversible rejection, or the incidence of rejection episodes. At 3 years after surgery, the estimated proportion of eyes with graft failure was 41% for the ABO-incompatible group and 31% for the ABO-compatible group (relative risk, 1.43; 95% confidence interval, 1.00 to 2.06). The estimated proportion of eyes with failure from rejection by 3 years was 30% for the ABO-incompatible group and 16% for the ABO-compatible group (relative risk, 1.98; 95% confidence interval, 1.25 to 3.13). The positive group in the Crossmatch Study had fewer graft failures, rejection failures, and rejection episodes than the negative group; however, these differences were not statistically significant. These studies demonstrate that, for high-risk patients who are immunosuppressed by topical steroid therapy and followed up according to the CCTS protocol: (1) neither HLA-A, -B nor HLA-DR antigen matching substantially reduces the likelihood of corneal graft failure; (2) a positive donor-recipient crossmatch does not dramatically increase the risk of corneal graft failure; and (3) ABO blood group matching, which can be achieved with relatively little effort and expense, may be effective in reducing the risk of graft failure.