The hormones glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) are secreted by enteroendocrine cells in the intestinal mucosa in response to nutrient ingestion. They are called incretin hormones because of their ability to enhance insulin secretion. However, in recent years it has become clear that the incretin hormones also affect glucagon secretion. While GLP-1 decreases glucagon levels, the effect of GIP on glucagon levels has been unclear. The regulation of glucagon secretion is interesting, as the combination of inadequate insulin secretion and excessive glucagon secretion are essential contributors to the hyperglycaemia that characterise patients with type 2 diabetes. Moreover, the near absence of a well-timed glucagon response contributes to an increased risk of hypoglycaemia in patients with type 1 diabetes. The overall aim of this PhD thesis was to investigate how the blood glucose level affects the glucagon and insulin responses to GIP in healthy subjects (Study 1) and patients with Type 2 diabetes (Study 2), and more specifically to investigate the effects of GIP and GLP-1 at low blood glucose in patients with Type 1 diabetes without endogenous insulin secretion (Study 3). The investigations in the three mentioned study populations have been described in three original articles. The employed study designs were in randomised, placebo-controlled, crossover set-up, in which the same research subject is subjected to several study days thereby acting as his own control. Interventions were intravenous administration of hormones GIP, GLP-1 and placebo (saline) during different blood glucose levels maintained (clamped) at a certain level. The end-points were plasma concentrations of glucagon and insulin as well as the amount of glucose used to clamp the blood glucose levels. In Study 3, we also used stable glucose isotopes to estimate the endogenous glucose production and assessed symptoms and cognitive function during hypoglycaemia. The results from the three studies indicate that GIP has effects on insulin and glucagon responses highly dependent upon the blood glucose levels. At fasting glycaemia and lower levels of glycaemia, GIP acts to increase glucagon with little effect on insulin release. At hyperglycaemia the insulin releasing effect of GIP prevail, which lead to an increase in glucose disposal by approximately 75% in healthy subjects (Study 1) and 25% in patients with Type 2 diabetes (Study 2) relative to placebo. After insulin-induced hypoglycaemia in patients with type 1 diabetes (Study 3), GIP increase glucagon release, which probably augments endogenous glucose production. This was associated with a reduced need for exogenously added glucose to prevent hypoglycaemia. In conclusion, the studies position GIP as a bifunctional blood glucose stabilising hormone that glucose-dependently regulates insulin and glucagon responses in humans.