To study the vasomotility of normal and diseased coronary arteries during dynamic exercise, symptom-limited supine bicycle exercise during cardiac catheterization was performed by 18 patients with classic angina pectoris. The cardiovascular response was assessed by hemodynamic measurements and computer-assisted determination of normal and stenotic coronary artery luminal areas from biplane coronary angiograms made before, during, and after exercise. After baseline measurements were recorded, 12 patients (group 1) performed bicycle exercise for 3.4 min (mean), reaching a maximum workload of 81 W (mean); at the end of exercise they received 1.6 mg sublingual nitroglycerin. After measurements at rest in six other patients (group 2), 0.1 mg intracoronary nitroglycerin was given, followed by exercise (3.8 min, 96 W; NS) and sublingual nitroglycerin as in group 1. During exercise in group 1, luminal area of the coronary stenosis decreased to 71% of resting levels (p less than .001), while area of the normal coronary artery increased to 123% of control (p less than .001). After sublingual nitroglycerin at the end of exercise, area of the normal vessel further increased to 140% of control (p less than .001), while luminal area of the stenosis dilated to 112% of resting levels (p less than .001 vs exercise, NS vs rest). Pretreatment with intracoronary nitroglycerin increased both normal (121%; p less than .05) and stenotic (122%; p less than .05) luminal areas, while preventing the previously observed narrowing of stenosis during exercise (114%; NS). Exercise resulted in a similar heart rate-systolic pressure product and caused angina pectoris in two-thirds of the patients in each group. However, patients pretreated with intracoronary nitroglycerin (group 2) had a lower mean pulmonary arterial pressure during maximum exercise (35 mm Hg) than those patients (group 1) not receiving pretreatment (47 mm Hg; p less than .001). Group 2 patients reached a percentage of their predicted work capacity (65%) that was about the same as that during previous upright bicycle exercise (71%; NS), while group 1 patients had a significantly lower work capacity (51% of predicted) than that before catheterization (82%; p less than .001). Hence, narrowing of coronary artery stenosis during dynamic exercise is attributable to active vasoconstriction due to its reversibility by preexercise intracoronary nitroglycerin. Patients who did not experience narrowing of stenosis during exercise (group 2) had less evidence of myocardial ischemia (lower mean pulmonary arterial pressure) and maintained their work capacity.(ABSTRACT TRUNCATED AT 400 WORDS)