The large and variable hepatic extraction of insulin is a major obstacle to our ability to quantitate insulin secretion accurately in human subjects. The evidence that C-peptide is secreted from the beta cell in equimolar concentration with insulin, but not extracted by the liver to any significant degree, has provided a firm scientific basis for the use of peripheral C-peptide concentrations as a semiquantitative marker of beta cell secretory activity in a variety of clinical situations. Thus, plasma C-peptide has proved to be extremely valuable in the study of the natural history of type 1 diabetes, to monitor insulin secretion in patients with insulin antibodies, and as an adjunct in the investigation of patients with hypoglycemic disorders. The use of the peripheral C-peptide concentration to accurately quantitate the rate of insulin secretion is more controversial. This is mainly because understanding of the kinetics and metabolism of C-peptide under different conditions is incomplete. Unfortunately, sufficient quantities of human C-peptide are not available to allow the experimental validation of the mathematical formulae that have been proposed for the calculation of insulin secretion from peripheral C-peptide concentrations. Until it is possible to perform such experiments, the accuracy of studies that have derived insulin secretion rates from peripheral C-peptide levels will remain uncertain. The assumption that the peripheral C-peptide:insulin molar ratio can be used as a reflection of hepatic insulin extraction has not been experimentally validated. The marked difference in the plasma half-lives of insulin and C-peptide complicates the interpretation of changes in their ratios.(ABSTRACT TRUNCATED AT 250 WORDS)