Aims/hypothesis. Insulin enhances coronary vasodilation in healthy subjects. We tested whether insulin is able to induce coronary vasodilation in Type I (insulin-dependent) diabetic mellitus patients. Additionally, the effect of short-term hyperglycaemia on myocardial perfusion was studied. Methods. Myocardial blood flow was quantitated basally and during adenosine infusion (140 µg/kg per min iv) with or without simultaneous insulin infusion (1 mU/kg per min for 60 min) in nine non-smoking Type I diabetic males (HbA 1c 7.4±1.0%) without diabetic complications and 10 healthy non-diabetic otherwise matched males using positron emission tomography and 15 O-water. Diabetic patients were studied on two occasions, once during normoglycaemia (plasma glucose ~6 mmol/l) and once during hyperglycaemia (~10 mmol/l) induced by reducing the dose of insulin for two days. Results. Resting myocardial blood flow was similar in the studied groups (NS). Hyperaemic adenosine stimulated flow was 23% lower in diabetic than in nondiabetic subjects (3.09±0.72 vs 4.0±1.13 ml·g -1 ·min -1 , p<0.05). Insulin increased significantly adenosine stimulated flow by 23% in diabetic and 17% in nondiabetic subjects (NS between the groups). Hyperglycaemia for two days had no effect on flow values when compared to the values during normoglycaemia (NS). Conclusion/interpretation. Insulin has similar vasodilative effects on coronary arteries in diabetic and nondiabetic subjects. Short-term hyperglycaemia does not alter myocardial blood flow or abolish insulin induced vasodilation in these patients. Insulin induced coronary vasodilation might contribute to the known beneficial effect of intensive insulin therapy on myocardial ischaemia in diabetic patients. [Diabetologia (2002) 45:775-782] Keywords Type I diabetes, myocardial perfusion, coronary vasoreactivity, insulin, hyperglycaemia, positron emission tomography.