Pancreatogenic diabetes (PD), secondary either to chronic calcific pancreatitis or to pancreatectomy, is characterized by higher frequency of hypoglycemic events during insulin therapy in comparison with type I insulin-dependent diabetes (IDD). Not only glucagon deficiency, but an enhanced peripheral tissue sensitivity to insulin could account for this metabolic behavior. We investigated several facets of insulin action, e.g., tissue sensitivity to insulin, insulin binding to red cells, and insulin kinetics in seven patients with PD in comparison with type I. Tissue sensitivity to insulin was evaluated by means of the glucose-insulin clamp technique as M/I x 100 ratio (mg . kg .-1 min-1/muU . ml-1), where M is the amount of glucose infused by Biostator GCIIS to clamp BG at basal level and I is the free insulin plateau concentration achieved by a primed-constant insulin infusion. At high BG 15 h after the last injection of regular insulin M/I x 100 was 7.79 (range 4.25-9.75) in PD and 4.20 (range 1.20-6.91) in D (P less than 0.05). At low and equal BG M/I x 100 was 8.55 (range 6.35-9.72) in PD and 3.42 (range 1.19-6.75) in D (P less than 0.01). The rate of endogenous glucose production was nearly totally suppressed in both groups of patients. Just before the two clamps, 125I-insulin specific binding to red cells was studied. The maximum specific binding was significantly higher in PD than in D at high BG (10.7 +/- 1.7 vs. 7.4 +/- 0.8/10(9) red cells) and at low and equal BG (12.4 +/- 1.2 vs. 6.8 +/- 0.8). Receptor concentration also was significantly higher in PD thant in D (P less than 0.02) while no significant differences were found in high affinity (Ke). Insulin kinetic data were analysed by using both "Model independent" (or noncompartmental) method and compartmental modeling. Patients with PD had significantly higher (P less than 0.05) plasma clearance of insulin.