Among 112 patients with sustained ventricular tachycardia, 15 were found to have exercise-induced symptomatic ventricular tachycardia. This population was divided into two subgroups: group 1A included five patients with coronary artery disease and group 1B consisted of 10 patients with no structural heart disease. All patients underwent clinical examination, exercise electrocardiography, left ventriculography, coronary angiography (n = 14) and electrophysiologic study. In group 1B, right ventriculography (n = 7), M mode and two-dimensional echocardiography were also obtained. Group 1A patients were compared with a population of 27 patients with coronary artery disease and chronic sustained ventricular tachycardia not related to exercise (group 2). There were no statistically significant differences between group 1A and group 2 in terms of age, sex, incidence of prior myocardial infarction, NYHA functional class, angina pectoris, symptoms during arrhythmia, severity and extent of coronary arterial lesions, ventricular dysfunction and wall motion abnormalities. In group 1B, coronary angiography and right and left ventricular function were normal. During electrophysiologic study, ventricular tachycardia was initiated in four group 1A patients. In group 1B, ventricular tachycardia was initiated in eight patients. In four of these patients ventricular pacing had to be combined with isoproterenol administration. In group 2, ventricular tachycardia was induced in 26/27 patients. From this study we conclude that in patients with coronary artery disease the electrophysiologic substrate of exercise-related sustained ventricular tachycardia does not differ from the substrate of non-exercise-related ventricular tachycardia. Re-entry is the most likely electrophysiologic mechanism. In patients without structural heart disease, the mechanism of the arrhythmia remains speculative.