In health, insulin is secreted in discrete pulses into the portal vein, and the regulation of the rate of insulin secretion is accomplished by modulation of insulin pulse mass. Several lines of evidence suggest that the pattern of insulin delivery by the pancreas determines hepatic insulin clearance. In previous large animal studies, the amplitude of insulin pulses was related to the extent of insulin clearance. In humans (and in large animals), the amplitude of insulin oscillations is approximately 100-fold higher in the portal vein than in the systemic circulation, despite only a fivefold dilution, implying preferential hepatic extraction of insulin pulses. In the present study, by direct hepatic vein sampling in healthy humans, we sought to establish the extent of first-pass hepatic insulin extraction and to determine whether the pattern of insulin secretion (insulin pulse mass and amplitude) dictates the hepatic insulin clearance and thereby delivery of insulin to extrahepatic insulin-responsive tissues. Five nondiabetic subjects (two men and three women, mean age 32 years [range 25-39], BMI 24.9 kg/m(2) [21.2-27.1]) participated. Insulin and C-peptide delivery from the splanchnic bed was measured in basal overnight-fasted state and during a glucose infusion of 2 mg . kg(-1) . min(-1) by simultaneous sampling from the hepatic vein and an arterialized vein along with direct estimation of splanchnic blood flow. Fractional insulin extraction was calculated from the difference between the C-peptide and insulin delivery rates from the liver. The time patterns of insulin concentrations and hepatic insulin clearance were analyzed by deconvolution and Cluster analysis, respectively. Cross-correlation analysis was used to relate C-peptide secretion and insulin clearance. Glucose infusion increased peripheral glucose concentrations from 5.4 +/- 0.1 to 6.4 +/- 0.4 mmol/l (P < 0.05). Likewise, insulin and C-peptide concentrations increased during glucose infusion (P < 0.05). Hepatic insulin clearance increased with glucose infusion (1.06 +/- 0.18 vs. 2.55 +/- 0.38 pmol . kg(-1) . min(-1); P < 0.01), but fractional hepatic insulin clearance was stable (78.2 +/- 4.4 vs. 84 0. +/- 3.9%, respectively; P = 0.18). Insulin secretory-burst mass rose during glucose infusion (P < 0.05), whereas the interburst interval remained unchanged (4.4 +/- 0.2 vs. 4.5 +/- 0.3 min; P = 0.36). Cluster analysis identified an oscillatory pattern in insulin clearance, with peaks occurring approximately every 5 min. Cross-correlation analysis between prehepatic C-peptide secretion and hepatic insulin clearance demonstrated a significant positive association without detectable (<1 min) time lag. Insulin secretory-burst mass strongly predicted insulin clearance (r = 0.81, P = 0.0043). In conclusion, in humans, approximately 80% of insulin is extracted during the first liver passage. The liver rapidly responds to fluctuations in insulin secretion, preferentially extracting insulin delivered in pulses. The mass (and therefore amplitude) of insulin pulses traversing the liver is the predominant determinant of hepatic insulin clearance. Therefore, through this means, the pulse mass of insulin release dictates both hepatic (directly) as well as extra-hepatic (indirectly) insulin delivery. These findings emphasize the dual role of the liver and pancreas and their relationship mediated through magnitude of insulin pulse mass in regulating the quantity and pattern of systemic insulin delivery.