This study examined the role of electrophysiologic study in the evaluation of patients with unexplained syncope. The incidence of abnormalities (conduction disease, supraventricular tachycardia, ventricular tachycardia) was compared in patient groups with and without heart disease, and the effect of treatment of these abnormalities on recurrence of syncope was examined. Electrophysiologic study was performed in 111 consecutive patients with syncope, with antiarrhythmic medications being discontinued 1 week prior to study. There was no mortality associated with the procedure and only 9 patients (8%) had complications (groin haematoma, atrial fibrillation or brady-arrhythmias persisting after the procedure). Abnormalities were detected in 31 of 73 patients with heart disease (42%) but in only 6 of 38 patients with no heart disease (16%, P less than 0.01). During follow-up, syncope recurred in 2 of 37 patients (5%) treated because of abnormal findings, compared with a recurrence rate of 24% (18 of 74 patients) in the untreated group (P less than 0.05). Probability of remaining free from syncope at 2 years was 0.94 in the treated group and 0.72 in the untreated group (P less than 0.05). Mortality during follow-up was confined to the heart disease group with 5 of 30 treated patients in this group dying (17%) compared with 3 of 43 untreated patients (7%, P = not significant). Syncope patients with heart disease were thus more likely to have a diagnostically useful study than patients with normal hearts. Treatment directed at correction of abnormalities detected at electrophysiologic study reduced recurrence of syncope but did not significantly affect mortality. Syncope did not appear to be a prelude for sudden death in patients with normal hearts. Electrophysiologic study had no mortality and low morbidity.