Acetylcholine-induced constriction of human coronary arteries in vivo is commonly attributed to endothelial dysfunction. To examine the effects of 2 other important determinants of vascular responses--namely, agonist concentration and the segment of circulation under study--the diameters of proximal, middle and distal segments of the left anterior descending artery (LAD) and coronary sinus oxygen saturation were measured in 10 patients with angiographically normal coronary arteries (group 1) and in 7 patients with coronary atherosclerosis (group 2) after intracoronary acetylcholine was infused at concentrations from 10(-7)M to between 10(-4)M and 10(-2)M. In group 1, acetylcholine caused minor (less than or equal to 6%) but progressive dilatation of the LAD up to 10(-4)M, but constriction, particularly of the distal segments and tertiary branches, occurred at higher concentrations. Over the same concentration range, coronary sinus oxygen saturation rose progressively from a basal level of 36 +/- 3% to a maximum of 72 +/- 3% in the absence of changes in heart rate and blood pressure, suggesting marked progressive dilatation of resistance vessels. Concentrations greater than or equal to 10(-3)M caused intense constriction of distal epicardial vessels and, in some cases, anginal pain and objective signs of ischemia. Conversely, in group 2, acetylcholine (infused only up to 10(-4)M for ethical reasons) failed to cause significant changes in LAD diameter. These data suggest that the local acetylcholine concentration and coronary vascular segment under study may determine the observed response to at least an equivalent extent as does the presence or absence of coronary atherosclerosis, raising the question of whether a constrictor response to intracoronary acetylcholine reliably indicates the presence of coronary atherosclerosis.