An oral gavage of either 3, 1 or 0.1 mmoles of 14C-labelled glucose was given to rats under standard feeding conditions or food deprived for 24 hr. The fate of the glucose label was determined at 10, 15, 30 and 60 min after gavage; at 60 min 40% of the glucose was absorbed in fed rats (60% in food deprived). The portal vein blood flows were determined and the levels of glucose, lactate, alanine and pyruvate, and their radioactivity, as well as that of CO2 were measured in both portal and arterial blood. The net computed glucose and 3-carbon carriers (lactate, alanine and pyruvate) actually released into the portal system by the intestine was lower than the amount of glucose taken up from the intestinal lumen in one hour. Oxidation to 14CO2 accounted for a 12-15% of the absorbed glucose. The size of the gavage deeply affected the proportion of glucose released into the portal blood (c. 50% with a 3 mmoles gavage and practically nil with a 0.1 mmoles gavage), but it affected much less the generation of lactate and other 3 C carriers. In fed rats, the net intestinal balance of non-radioactive glucose was negative, and that of lactate positive; when radioactive glucose was considered, the pattern was inverted. In starved rats, both glucose and lactate were released in large proportions by the intestine, but alanine efflux was lower. It can be concluded that the intestine consumes a considerable proportion of glucose in the fed state. Glucose handling by the intestine is compartmentalized in two functional circuits: glucose is taken up from the arterial blood and used for intestinal metabolism and lactate production, luminal glucose is absorbed mainly unaltered and transferred to the portal blood. Thus, the generation of lactate is mainly related to the availability of arterial glucose. In addition to the release of the ingested glucose as 3 C carriers or glucose, an extraportal pathway for glucose transfer into the bloodstream is postulated.