The small intestines of anaesthetized rats were perfused in situ for 60 min with 0.5% solutions of glucose, maltose, sucrose, maltodextrin DE 20, maltodextrin DE 5 or starch. Blood samples were repeatedly taken from the v. portae and the v. femoralis to estimate blood glucose and serum insulin levels as a function of perfusion time. The experiment was also performed to clarify whether a correlation exists between these parameters and the substrate uptake from the intestinal lumen (determined in the first part of the study). The highest glucose levels in v. portae and the highest portal-peripheral differences were found when glucose and maltose solutions were administered. Glucose levels in v. femoralis were almost independent of the substrate perfused. Perfusion with glucose caused a considerable insulin secretion starting immediately with the onset of perfusion. It was concluded that glucose might also stimulate insulin secretion on the intestinal level. Perfusion with maltose also effected a significant insulin output, the start of which was delayed however, compared with the effect of glucose.