The present study was performed to evaluate the porto-systemic insulin gradient in response to (1) glucose feeding (2) intramuscular insulin administration, and (3) peritoneal insulin administration in unanesthetized swine. The experiment was designed to verify the hypothesis that intraperitoneal insulin administered might lead to a more physiologic portal vein insulin concentration than systemic administration of a similar insulin dose. Studies were performed in 4 domestic swine with chronic, indwelling catheters in the inferior vena cava and portal vein. Unpaired studied of the absolute portal venous and systemic venous insulin concentrations were performed in response to glucose prn(n = 4), 1 unit regular insulin/kg i.m.(n = 4), and 1 unit regular insulin/kg i.p.(n = 5). Timed blood samples were obtained and serum insulin concentrations determined by RIA. Portal and caval serum insulin concentrations following intramuscular insulin injection showed no significant difference. A significant portal vein insulin excess (p less than 0.001) was demonstrated following both feeding and intraperitoneal insulin.