We measured transhepatic C-peptide and insulin concentrations in plasma, and hepatic removal of insulin, to examine whether the practice of reporting the C-peptide:insulin molar ratio as a measure of the hepatic removal of insulin is valid. In anesthetized dogs (n = 6), during electromagnetic hepatic blood flow monitoring, endogenous insulin was suppressed with somatostatin, while equimolar proportions of porcine insulin and simian C-peptide (2.4 and 6.0 pmol/kg.min) were infused during two consecutive 45-min periods. Insulin reached steady state within 20 min (t1/2 = 4.5 min); however, C-peptide concentrations continued to rise (t1/2 V 12.5 min). The ratio decreased when the peptide infusion was changed to the higher rate and increased when it was stopped, reflecting the more rapid removal of insulin than of C-peptide. Hepatic removal of insulin remained constant during the two infusion periods (average 60% extraction) and never correlated with the changing molar ratios. Hepatic net flux of insulin correlated with the ratio (P less than 0.05) only while plasma insulin concentrations were rising during constant-rate infusion. We therefore conclude that the molar ratio is not a reliable measure of the hepatic removal of insulin during non-steady states of insulin or C-peptide.