Islet allografts in insulin-dependent diabetic (IDDM) patients exhibit variable survival lengths and low rates of insulin-independence despite treatment with anti-T-cell antibodies and maintenance immunosuppression. Use of poorly characterized freshly isolated preparations makes it difficult to determine whether failures are caused by variations in donor tissue. This study assesses survival of standardized beta-cell allografts in C-peptide negative IDDM patients on maintenance immunosuppression following kidney transplantation and without receiving anti-T-cell antibodies or additional immunosuppression. Human islets were isolated from pancreatic segments after maximal 20 h cold-preservation. During culture, preparations were selected according to quality control tests and combined with grafts with standardized cell composition (> or = 50% beta cells), viability (> or = 90%), total beta-cell number (1 to 2 x 10(6)/kg body weight) and insulin-producing capacity (2 to 4 nmol x graft(-1) x h(-1)). Grafts were injected in a liver segment through the repermeabilized umbilical vein. After 2 weeks C-peptide positivity, four out of seven recipients became C-peptide negative; two of them were initially GAD65-antibody positive and exhibited a rise in titre during graft destruction. The other three patients remained C-peptide positive for more than 1 year, two of them becoming insulin-independent with near-normal fasting glycaemia and HbA1c; they remained GAD65- and islet cell antibody negative. The three patients with surviving grafts presented a history of anti-thymocyte globulin therapy at kidney transplantation. Long-term surviving grafts increased C-peptide release following intravenous glucagon or oral glucose but not following intravenous glucose. Thus, cultured human beta-cells can survive for more than 1 year in IDDM patients on maintenance anti-rejection therapy for a prior kidney graft and without the need for an increased immunosuppression at the time of implantation. The use of functionally standardized beta-cell grafts helps to identify recipient and graft factors which influence their survival and metabolic effects. Insulin-independence can be achieved by injection of 1.5 million beta-cells per kg body weight in a liver segment. These beta-cell implants respond well to adenylcyclase activators but poorly to glucose.