The absorption mechanism of insulin administered in the peritoneal cavity (IP) is of current interest because of the near availability of implantable insulin-infusion devices for the treatment of diabetes. To determine the fraction of insulin absorbed by the portal circulation after IP administration, a novel nonisotopic method is described. Conscious fasting diabetic dogs were studied at normoglycemia via the euglycemic insulin-clamp method. Posthepatic appearance of insulin and C-peptide were measured in peripheral blood during IP or intravenous (IV) equimolar infusion of insulin and C-peptide at two sequential 3-h infusion rates of 3.2 and 12.8 pmol.kg-1.min-1. Prior studies have shown that 40-60% of portal insulin is extracted at first pass by the liver, whereas C-peptide is not extracted. Thus, the fraction (F) of IP insulin not taken up by liver at first pass and consequently the fraction absorbed by the portal circulation can be derived from insulin (I) and C-peptide (C) plasma concentration values at steady state with a monocompartmental model where F = (IIP/IIV)(CIV/CIP). The mean +/- SE value of F was 49.7 +/- 8.8%. Glucose disappearance rates were lower with IP than IV infusion but similar when peripheral insulin levels were matched. We conclude that IP insulin is almost entirely absorbed by the portal circulation and induces lower glucose disappearance rates than IV insulin because of lower peripheral circulating insulin levels. Whether these properties make the IP route a more appropriate route for insulin therapy than the subcutaneous or IV routes remains to be established.