Total plasma immunoreactive pancreatic glucagon (IRG) was measured in samples taken simultaneously from the proximal portal vein and superior vena cava of 26 healthy rats. The portal-peripheral ratio of IRG was 2.80+/-0.25, the portal-peripheral difference (Delta) 124+/-15 pg/ml, and percentage extraction 58+/-3. Gel filtration of paired portal and peripheral vein samples showed that reduction in the 3,500-dalton IRG component (glucagon) in peripheral samples accounted for almost all the differences, there being minimal and inconsistent changes in the high molecular weight (>40,000) fraction. The portal-peripheral ratio of the 3,500-dalton glucagon was 5.24+/-1.10, the portal-peripheral difference 130+/-33 pg/ml, and the percentage extraction 81+/-5. To study the transhepatic differences in the 9,000-dalton "proglucagon-like" material, the experiment was repeated in nine rats 24 h after bilateral nephrectomy, a procedure which increases plasma levels of this fraction. The portal-peripheral ratio for plasma IRG in these rats was 1.48+/-0.12, the portal-peripheral difference 140+/-29 pg/ml, and percentage extraction 28+/-5. Gel filtration revealed no consistent differences between portal and peripheral concentrations of the 9,000- and >40,000-dalton components, which comprised 40 and 13%, respectively, of the mean IRG level of 492+/-35 pg/ml. In contrast, there were marked differences between portal and peripheral levels of the 3,500-dalton component the ratio being 3.42+/-0.63, the portal-peripheral difference 182+/-32 pg/ml, and percentage extraction 64+/-5. Similar studies in a healthy dog, in which species there are significant circulating levels of the 9,000-dalton IRG component, confirmed the selective hepatic extraction of the 3,500-dalton fraction. We conclude that the various IRG fractions are metabolized differently by the liver, and that portal-peripheral ratios based on direct assay of plasma IRG will vary depending on the percentage glucagon immunoreactivity in each fraction; the greater the combined contribution of fractions other than the 3,500-dalton component to total plasma IRG, the lower will be the ratio. Because of the heterogeneity of circulating IRG and significant differences in the metabolism of its various components, gel filtration of plasma samples is necessary for precise quantitation of the hepatic uptake of each particular fraction.