Postprandial symptoms, known as the dumping syndrome, often complicate total or partial gastrectomy and can be classified as ªearlyº or ªlateº dumping [1,2,3] according to the temporal relation to the preceding meal. Early symptoms consist of neurovascular and gastrointestinal symptoms appearing within 30 min after eating, while the late symptoms (sweating, paleness, fatigue, muscular weakness, blurred consciousness, anxiety, tremor, tachycardia) appear after 2 h or later and are related to hypoglycaemia. After test meals or oral glucose tolerance tests, gastrectomised patients show rapid gastric emptying and increased intestinal motility [4,5,6,7] which are thought to be responsible for the quick and steep rise in plasma glucose concentration. At first, the abnormal hyperglycaemic response itself was thought to be responsible for the pronounced concomitant hyperinsulinaemia seen in these patients [1] but similar plasma glucose concentrations obtained by i. v. glucose infusion do not induce equally high insulin and Diabetologia (1998) Summary The plasma concentrations of the insulinotropic incretin hormone, glucagon-like peptide-1 (GLP-1) are abnormally high after oral glucose in partially gastrectomised subjects with reactive hypoglycaemia, suggesting a causal relationship. Because of the glucose-dependency of its effects, it is impossible to induce hypoglycaemia in normal subjects in the basal state by exogenous GLP-1, regardless of dose. To further assess the role of the incretin hormones in reactive hypoglycaemia, we reproduced the glucose and hormone profiles of the patients with reactive hypoglycaemia in 8 healthy volunteers in 4 separate protocols: 1) i. v. infusion of glucose (25 g) alone, 2) glucose together with i. v. GLP-1 infusion, and 3) and 4) glucose together with i. v. infusion of the other incretin hormone, glucose-dependent insulinotropic polypeptide (GIP), at two different infusion rates. The plasma glucose, GLP-1 and GIP concentrations (low dose) obtained were comparable with those of the patients. With GLP-1, infusion of a total of 33.4 ± 1.3 g glucose was required to obtain plasma glucose concentrations similar to those obtained by glucose infusion alone; with low GIP, 28.0 ± 1.2 g and with high GIP 38.4 ± 3.5 g. Insulin concentrations increased 10-fold with GLP-1 compared with i. v. glucose alone, but less with high and low GIP. In contrast, C-peptide concentrations were similar after GLP-1 and high GIP. After termination of i. v. glucose the lowest glucose concentrations were 4.5 (3.7±4.9) (median, range) for glucose alone; 2.4 (1.9±2.8) mmol/l with GLP-1; 3.7 (2.6±4.0) with low GIP and 3.3 (2.1±4.2 ) with high GIP. Thus, the exaggerated GLP-1 response to nutrients in patients with accelerated gastric emptying could be responsible for their high incidence of postprandial reactive hypoglycaemia. [Diabetologia (1998)