Sepsis and extensive burn injury produce clinical syndromes characterized in part by "insulin resistance," but it is unclear if these insulin resistant states are identical. To test if the maximal biological effectiveness of insulin is altered in septic or burned patients, eight septic patients and eight nonseptic patients recovering from severe burn injury were studied using the hyperinsulinemic eukalemic euglycemic clamp technique. Compared with bed-rested controls, the septic patients showed an insulin-induced plasma clearance of potassium, which was 183% higher (P less than .001), and a concomitant glucose clearance, which was 52% lower (P less than .001). Nonseptic burn patients also had a 91% increase in potassium clearance (P less than .05), but their maximal insulin-stimulated glucose uptake was not different from that of bedrested controls. When septic patients were compared with their nonseptic burned counterparts, there was no difference in potassium clearance in response to insulin, but glucose uptake by the septic patients was 47% lower (P less than .001). Insulin infusion completely suppressed hepatic glucose production in both septic patients and in nonseptic burn patients. The percent of whole body glucose uptake that was oxidized was not different between the septic patients and the nonseptic postburn patients in both the basal and insulin-stimulated states (38% and 51% v 38% and 42%, respectively). It is concluded that septic and postburn insulin resistance differ in that peripheral glucose uptake in sepsis, but not nonseptic burn injury, is refractory to pharmacologic insulin stimulation, whereas in both states insulin effectively stimulates potassium uptake.