Since our report at the 1984 American Surgical Association meeting of 100 pancreas transplants from 1966 through 1983, another 190 have been performed. The current series, begun in 1978, now numbers 276 cases, and includes 133 nonuremic recipients of pancreas transplants alone (PTA), 46 simultaneous pancreas/kidney transplants (SPK), and 97 pancreas transplants after a kidney transplant (PAK). Duct management techniques used were free intraperitoneal drainage in 44 cases, duct occlusion in 44, enteric drainage in 89, and bladder drainage in 128. The 1-year patient and graft survival rates in the entire cohort of 276 were 91% and 42%. One-year patient survival rates were 88% in the first 100, 91% in the second 100, and 92% in the last 76 cases; corresponding 1-year graft survival rates were 28%, 47%, and 56% (p < 0.05). A prospective comparison of bladder drainage (n = 82) versus enteric drainage (n = 46) in PAK/PTA cases since November 1, 1984 favored bladder drainage (1-year graft survival rates of 52% vs. 41%) because of urinary amylase monitoring. The best results were in recipients of primary SPK bladder-drained transplants (n = 39), with a 1-year pancreas graft survival rate of 75%, kidney graft survival rate of 80%, and patient survival rate of 95%. Logistic regression analysis, with 1-year graft function as the independent variable, showed significant (p < 0.05) predictors of success (odds ratio) to be technique: bladder drainage (5.8) versus enteric drainage (2.5) versus duct injection (1.0); category: SPK (6.0) versus PAK from same donor (3.2) versus PAK from different donor (1.2) versus PTA (1.0); and donor HLA DR mismatch: 0 (5.0) versus 1 (2.5) versus 2 (1.0) antigens. On April 1, 1989, 90 patients had functioning grafts (60 euglycemic and insulin-free for more than 1 year, 10 for 5 to 10 years); these, along with 24 others whose grafts functioned for 1 to 6 years before failing, are part of an expanding cohort in whom the influence of inducing a euglycemic state on preexisting secondary complications of diabetes is being studied. Only preliminary data is available. In regard to neuropathy, at more than 1 year after transplant in patients with functioning grafts, conduction velocities in some nerves were increased over baseline. In regard to retinopathy, deterioration in grade occurred in approximately 30% of the recipients by 3 years, whether the graft functioned continuously or failed early, but thereafter retinopathy in the patients with functioning grafts remained stable. In patients with functioning pancreas grafts, kidney biopsies have shown a decreased glomerular mesangial volume compared to diabetic controls. Pancreas transplantation is increasingly successful in both uremic and nonuremic diabetic patients, and may ameliorate secondary complications of diabetes.