The endothelium plays an important role in the regulation of vascular tone. Although animal data show evidence for an impaired endothelium-dependent vasodilation in diabetes, human in vivo data are scarce. We investigated 11 type I diabetic patients and 11 matched healthy control subjects. The diabetic patients were selected on their relatively poor metabolic regulation (HbA1c > 8.5%), but none showed signs of microvascular complications. In all subjects, we recorded the forearm vasodilator responses to three different stimuli: 1) 5 min of forearm ischemia to obtain a maximal vasodilator response; 2) infusion of MCh into the brachial artery (dosages: 0.03-0.3-1.0 micrograms.min-1.100 ml-1 forearm volume) to evaluate endothelium-dependent vasodilation; and 3) intra-arterial infusion of SNP (dosages: 0.06-0.2-0.6 micrograms.min-1.100 ml-1) to evaluate endothelium-independent vasodilation. The diabetic patients had their usual subcutaneous insulin dose and breakfast 90 min before the start of the test. Baseline levels of BP and FBF were similar in both groups. The PORH response was similar in both groups, with a percentage fall in FVR of 92 +/- 1% in diabetic patients and 94 +/- 1% in control subjects. In the control subjects, MCh infusions exerted a dose-dependent vasodilator response with a maximal fall in the FVR of 90 +/- 2%. The highest dose of SNP induced a fall in FVR of 81 +/- 6% in this group. In diabetic patients, the percentage decrements in FVR during the several dosages of MCh and SNP were similar when compared with the control group. We conclude that chronic hyperglycemia, as occurred in our patients with uncomplicated diabetes mellitus, does not impair endothelium-dependent vasodilation in vivo.