The relative tissue specificities of biosynthetic human proinsulin and porcine insulin were compared in eight insulin-deficient diabetic patients. Variable proinsulin and insulin infusions were administered iv for at least 18 h on separate occasions to achieve euglycemia. Once equivalence with respect to this end point was established, the paired infusions were made constant and compared with respect to their effect on plasma glucose, hepatic glucose output (Ra), glucose utilization, glucose clearance, and blood beta-hydroxybutyrate, alanine, and glycerol in response to a glucagon infusion (3 ng/kg X min). Basal Ra did not differ between euglycemia established using proinsulin vs. insulin [1.92 +/- 0.12 vs. 1.65 +/- 0.07 mg/kg X min] when data were analyzed by multivariate analysis of variance; likewise, the increment in Ra in response to glucagon infusion did not differ between the two treatments (1.26 +/- 0.22 vs. 1.17 +/- 0.24 mg/kg X min). Blood beta OHB, alanine, and glycerol also did not differ between the insulin and proinsulin treatments. Under these conditions of prolonged infusion in which the differences in clearance properties between proinsulin and insulin do not play a role and which maintain basal euglycemia, the two compounds equally limit the response to a stimulus to hepatic glucose production. Under nonsteady state conditions, when the differences in clearance rates are operative, proinsulin may have relatively greater effects on Ra than do doses of insulin that are biologically equivalent after prolonged infusions.