We evaluated the influence of donor-recipient HLA compatibility and recipient pretransplant antidonor sensitization on liver allograft recipient survival. The overall graft survival results for 67 cyclosporine-prednisone treated liver allograft recipients at 3, 6, and 12 months posttransplant were 86%, 83%, and 83%, respectively. No significant differences were observed when comparing the one-year survivals of 81% vs. 85% for men and women or 80% vs. 85% for adult and pedi-atric patients. Similarly, no differences were observed when comparing one-year graft survivals for well vs. poorly matched recipients of 77% vs. 83% for recipients with <2 HLA A, B vs. >2 HLA A, B mismatches (MMs) and 82% vs. 82% for recipients with 0–1 HLA-DR MMs vs. 2 HLA-DR MMs, respectively. Pretransplant transfusion history and race also did not influence survival. Standard NTH (long-incubation) and anti-human globulin (AHG) crossmatches were performed. The 12% of recipients (8/67) displaying a positive NIH crossmatch experienced significantly poorer 3-, 6-, and 12-month survivals of 62% vs. 89%, 62% vs. 86%, and 62% vs. 86% (all P<0.05), respectively, than the 59 NIH-crossmatch negative recipients. Similarly, the 16% (11/ 67) of recipients displaying a positive AHG crossmatch had significantly poorer 3-, 6-, and 12-month survivals of 63% vs. 91%, 54% vs. 89%, and 54% vs. 89% (all P<0.05) respectively, than the 56 AHG crossmatch-negative recipients. NIH and AHG crossmatch-positive sera were treated with dithioerythritol (DTE) to establish whether reactivity was due to IgM or IgG immunoglobulin. One-year graft survivals of 65% vs. 30% (P<0.05) were observed when the crossmatch-positive sera reactivities were due to IgM vs. IgG immuno-globulin. While graft survivals were improved when positive crossmatch serum reactivity was due to IgM, these survivals were still significantly poorer than when the crossmatches were completely negative (86% vs. 60%, P<0.05 for NIH-negative vs. NIH-positive, but DTE-negative, and 88% vs. 77%, P<0.05 for AHG-nega-tive vs. AHG-positive, but DTE-negative). Therefore, an NIH- or AHG-positive crossmatch, due either to IgM or IgG reactivity, results in poor early (3− and 6-months) liver allograft survival. Crossmatch-positive recipients experienced significantly (P<0.05) more rejections and more steroid-resistant rejections (P<0.05) than crossmatch-negative recipients. Of great interest was the fact that 55% (6/11) of recipients displaying a positive pretransplant crossmatch experienced a moderate initial rejection episode compared with only 9% (5/56) for crossmatch-negative recipients (P<0.05). These results suggest, therefore, that a positive crossmatch adversely affects the early (3− and 6-month) survival of primary liver allografts. Since a crossmatch often cannot be performed prior to the transplant operation, the clinician's knowledge of the crossmatch results could be of importance in the choice of immunosuppressive therapy for crossmatch-positive recipients during the early posttransplant period.