A means of estimating human enteroglucagon (glucagon-like immunoreactivity of intestinal origin) in tissues and plasma is described, based on the subtraction of RIA values obtained with the C-terminal-directed glucagon antiserum RCS5 from the total glucagon-like immunoreactivity determined with the N-terminal- to midmolecule-directed glucagon antiserum R59. Gel filtration on Sephadex G-50 of human plasma and extracts of normal human intestine separated the R59 immunoreactivity into three peaks: a small peak of void volume material, a major peak coeluting with porcine glicentin, and a smaller peak coeluting with pancreatic glucagon. No RCS5 immunoreactivity was detected in the human gut, except for a small amount constituting less than 2% of the total glucagon-like immunoreactivity in the ileum and rectum only. In extracts of human pancreas, the chromatographic profiles obtained with RCS5 and R59 assays differed from the intestinal patterns, but were identical to each other, giving no evidence of a significant amount of pancreatic R59 immunoreactivity that was not also reactive with RCS5. Chromatography of plasmas from healthy subjects and patients with dumping syndrome, active coeliac disease, and tropical sprue showed that only the second major peak of R59 immunoreactivity reflected the basal or postnutrient increases in the plasma enteroglucagon concentration. In patients with exaggerated enteroglucagon release, the rise was again found to be entirely due to an increase in this peak of immunoreactivity. This major molecular form of human enteroglucagon, similar in size to porcine glicentin, is, thus, the form most likely to be of physiological and pathophysiological significance.