335 Preformed HLA antibodies have been associated with poor kidney allograft survival and the development of HLA alloantibodies post transplant has also been associated with rejection and subsequent graft loss. The purpose of this study was to evaluate the predictive nature of HLA antibody testing in identifying high risk patients receiving a pancreas transplants and to determine whether antibody detected by post transplant monitoring was associated with rejection events and graft outcomes. The group of patients selected were 150 pancreas transplants were performed between 6/18/90-9/11/96 for which we had stored serum and follow up > 1 yr.. 133 received SPK, 5 LRD kidney/CAD pancreas, and 12 pancreas alone. All were transplanted with negative AHG T cell crossmatches: 131 with negative T and B flow crossmatches (FXM). The median age was 35 yrs. & 43% were female. The mean HLA mismatches were 3.9 for ABDR and 1.9 for DRDQ. Kidneys and pancreata were transplanted using bladder drainage. Patients received induction OKT3 therapy with maintenance immunosuppression consisting of cyclosporine, Imuran and prednisone. The outcomes for this cohort were the following: pancreas, kidney and patient survival at one year were 86.7%, & 88.7% and 92.7%. The overall rejection rate was 86.7% as determined by protocol and dysfunction biopsies. The average rejection reoccurrence rate was 80%, mean number of rejections was 2.29, and average time to first rejection was 44.5 days. The presence of alloantibody in pre-transplant sera was determined using PRA-STAT and CDC-PRA; the presence of alloantibody specific for donor was determined using FXM techniques. Serum collected post-transplant was tested at 1,2,3,6, & 12, >12 mo. using PRA-STAT (82 patients) and FXM (78 patients). The outcomes evaluated included kidney and pancreas loss up to 1, 3, 6 and 12 mo., patient survival at 12 mo, the total number and severity of rejections, time to first rejection, and rejection reoccurrence. PRA was ranked at 5, 7, 10, 15, 20, 30 and 50% and PRA evolution was an increase of more than 9%. Alloantibody in pre-transplant serum detected by PRA-STAT, CDC-PRA or FXM was associated with higher incidence of early pancreas rejection and lower pancreas graft survival at one year and kidney loss in low risk patients at 120 days. PRA stat evolution was associated with pancreas graft loss. Alloantibody detected by FXM in post-transplant serum was highly correlated (p<0.01) with a pancreas rejection event in the first 6 weeks after transplant and with both pancreas and kidney rejection events occurring between 1 and 6 months post transplant (p<0.001). This correlation was not seen with PRA-STAT. Conclusions: 1) Alloantibody detected pre-transplant identifies patients at risk for early pancreas rejection and lower pancreas graft survival. 2) Alloantibody is generated post-transplant during pancreas and kidney rejection events and may be used to guide biopsy 3) Flow crossmatching appears to be more sensitive than PRA-STAT in detecting antibody after transplant.