Aims/hypothesis. Treatment with intravenous glucose-insulin-potassium has beneficial effects in reperfused patients, reducing mortality in patients with myocardial infarction by 28%. We hypothesized that insulin response to glucose-insulin-potassium infusion might lead to vasodilation in ischemia/reperfusion (I/R). Hyperglycaemia and hyperinsulinaemia determine oxidative stress. We therefore investigated the microcirculatory changes following I/R after glucose-insulin-potassium or in association with glucose-insulin-potassium dipyridamole in hamster cheek pouch. Methods. The control (I/R), glucose-insulin-potassium groups with and without dipyridamole were treated with saline, 300 g/l, 50 U/l insulin and 80 meq/l KCl infused at 0.2 ml (.) 100 g(-1) (.) h(-1), and GIK plus dipyridamole (0.084 mg (.) 100 g(-1) intravenously) at beginning, 30 min before ischaemia, and continuing through reperfusion. We measured microvessel diameter changes, arteriolar red blood cell velocity, permeability increase, capillary perfused length, leukocyte and platelet adhesion. Results. Hyperglycaemia and hyperinsulinaemia did not cause vasodilation whereas in the glucose-insulin-potassium group with dipyridamole there was a marked arterial vasodilation with increased red blood cell velocity and perfused capillary length at reperfusion. Glucose-insulin-potassium infusion reversed the arterial vasoconstriction caused by I/R at reperfusion. Adhering leukocytes to venules decreased by 56 and 86% while platelets adhering to microvessels was reduced by 52 and 72% at reperfusion in glucose-insulin-potassium groups with and without dipyridamole, respectively. The permeability was decreased by GIK and completely suppressed by GIKD after I/R. Conclusion hypothesis. We demonstrated that GIK, when used in combination with dipyridamole, had beneficial effects on the capillary perfusion against I/R-induced injury. There was a marked reduction of leukocyte and platelet adhesion that can be explained by the antioxidant properties of dipyridamole.