The effects of fentanyl and sufentanil with and without N2O on left ventricular myocardium supplied by a critically narrowed and a normal coronary artery were studied in 16 dogs. Regional ventricular function was measured by recording ventricular segment length with the use of ultrasonic length detectors in the left anterior descending (LAD) and the left circumflex (LC) coronary artery territories before and during critical stenosis of the LAD. Critical stenosis was documented by the absence of a hyperemic response following a 10-s total occlusion of the LAD. Hemodynamic variables (aortic flow and pressure, left ventricular pressure, heart rate, and coronary blood flow) were measured and the first derivative of left ventricular pressure (LVdP/dt) and coronary perfusion pressure derived. Eight dogs received fentanyl 100 μg · kg−1followed by an infusion of 1 μg · kg−1· min−1while ventilated with O2:N2(1:2), and eight dogs received sufentanil 30 μg · kg−1with an infusion of 0.3 μg · kg−1· min−1. Replacement of N2with N2O produced evidence of mild systolic myocardial depression but no dysfunction in either group. After application of the critical constriction, the addition of N2O rapidly produced evidence of dysfunction with significant postsystolic shortening only in the LAD territory. This was not accompanied by hypotension or a decrease in coronary flow and was not always reversible. Higher infusion rates of either narcotic (fentanyl 2 μg · kg−1· min−1, 4 μg · kg−1· min−1; sufentanil 0.6 μg · kg−1· min−1, 1.2 μg · kg−1· min−1) in the absence of N2O did not produce dysfunction but had no protective effect when N2O was added. These data suggest that the addition of N2O to a narcotic base anesthetic can produce clinically unapparent regional myocardial ischemia, which may be an important factor in the etiology of perioperative myocardial infarction.