OBJECTIVE To evaluate, on the basis of long-term outcome, the value of electrophysiologic exploration for determining the cause of syncope of unknown origin after clinical examination and standard laboratory tests. METHODS In 1985 and 1986, 109 consecutive patients with syncope of unknown origin underwent electrophysiologic exploration. The patients were divided into two groups. Baseline electrocardiogram was abnormal in 59 patients: cardiopathy (32%), ischemia (17%). In 44 of these patients, electrophysiologic exploration revealed major abnormalities requiring treatment (Group 1). No aetiology could be defined in 65 patients (Group 2). Patients were followed for 37 +/- 11 months. Three were lost to follow-up and 106 were retained for analysis. RESULTS Morbidity and mortality were 25% and 16% respectively at the end of follow-up. Two sudden deaths were recorded, both in Group 1. Total mortality was greater in Group 1 (10 vs 6) so electrophysiologic exploration had defined a population at lower risk of sudden death and overall mortality. Relapse of syncope was observed in 27 patients (25% and 26% in Groups 1 and 2 respectively) due to rhythm disorders in 8, vasodepression in 1 and unknown origin in 18. CONCLUSIONS The rate of recurrent syncope suggests these patients should have a second work-up in order to diagnose initially false negatives. Relapse raises the problem of electrophysiologic abnormalities and multifactorial mechanisms causing syncope of unknown origin.