Acute myocardial ischemia decreases conduction velocity through the ischemic myocardium and may produce QRS prolongation on the surface electrocardiogram. To investigate the effect of exercise-induced myocardial ischemia on the QRS duration, 124 patients who underwent diagnostic cardiac catheterization and exercise thallium-201 scintigraphy were studied. Seventeen patients had normal coronary arteries and 107 had coronary artery disease. QRS duration decreased with exercise in patients with normal coronary arteries (-3.7 ms, confidence interval -7.7 to 0.3) but increased in patients with coronary artery disease; exercise-induced QRS prolongation was directly related to the number of vessel disease (1.2 ms in patients with one-vessel disease; 8.7 ms in patients with two-vessel disease, and 10.5 ms in patients with three-vessel disease, p < 0.001). Likewise, exercise-induced QRS duration decreased in patients without exercise-induced reversible perfusion defects (-1.3 ms, confidence limits -2.8 to 0.2) but increased in patients with exercise-induced perfusion defects (9.7 ms in patients with one perfusion defect; 13.6 ms in patients with two perfusion defects, and 16.3 ms in patients with three perfusion defects, p < 0.001). Exercise-induced QRS prolongation was best related to the number of exercise-induced perfusion defects than to the number of vessel disease, p < 0.001. In conclusion, exercise-induced QRS prolongation in patients with coronary artery disease was directly related to the number of vessel disease and to exercise-induced perfusion defects. This QRS prolongation was better related to the number of reversible perfusion defects than to the number of vessel disease. Thus, exercise-induced QRS prolongation in patients with coronary artery disease may be a marker of exercise-induced myocardial ischemia.