Previous studies showed that nitric oxide (NO) plays an important role in coronary arteriolar dilation to increases in myocardial oxygen consumption (MVO(2)). We sought to evaluate coronary microvascular responses to endothelium-dependent and to endothelium-independent vasodilators in an in vivo model. Microvascular diameters were measured using intravital microscopy in 10 normal (N) and 9 hyperglycemic (HG; 1 wk alloxan, 60 mg/kg iv) dogs during suffusion of acetylcholine (1, 10, and 100 microM) or nitroprusside (1, 10, and 100 microM) to test the effects on endothelium-dependent and -independent dilation. During administration of acetylcholine, coronary arteriolar dilation was impaired in HG, but was normal during administration of nitroprusside. To examine a physiologically important vasomotor response, 10 N and 7 HG control, 5 HG and 5 N during superoxide dismutase (SOD), and 5 HG and 4 N after SQ29,548 (SQ; thromboxane A(2)/prostaglandin H(2) receptor antagonist) dogs were studied at three levels of MVO(2): at rest, during dobutamine (DOB; 10 microg. kg(-1). min(-1) iv), and during DOB with rapid atrial pacing (RAP; 280 +/- 10 beats/min). During dobutamine, coronary arterioles dilated similarly in all groups, and the increase in MVO(2) was similar among the groups. However, during the greater metabolic stimulus (DOB+RAP), coronary arterioles in N dilated (36 +/- 4% change from diameter at rest) significantly more than HG (16 +/- 3%, P < 0.05). In HG+SQ and in HG+SOD, coronary arterioles dilated similarly to N, and greater than HG (P < 0.05). MVO(2) during DOB+RAP was similar among groups. Normal dogs treated with SOD and SQ29,548 were not different from untreated N dogs. Thus, in HG dogs, dilation of coronary arterioles is selectively impaired in response to administration of the endothelium-dependent vasodilator acetylcholine and during increases in MVO(2).