Electrophysiologic testing was performed in 53 patients with recurrent syncope that remained unexplained despite a thorough neurologic and noninvasive cardiac evaluation. Fifteen patients had no structural heart disease, 9 had mitral valve prolapse and 29 had structural heart disease other than mitral valve prolapse. Nonsustained ventricular tachycardia was induced in 15 patients (28%), sustained ventricular tachycardia was induced in 9 (17%), ventricular fibrillation was induced in 4 (8%) and sinus node function was abnormal in 2 (4%). Female sex and lack of structural heart disease were independently associated with a negative electrophysiologic study (p less than 0.001). Patients with inducible ventricular tachycardia or ventricular fibrillation were treated with drugs selected on the basis of the results of electropharmacologic testing. The recurrence rate of syncope was 43% over a 31 +/- 10 month period (mean +/- standard deviation) of follow-up in patients with a negative electrophysiologic study, 40% over a 22 +/- 6 month period in patients with inducible nonsustained ventricular tachycardia, 0% over a 30 +/- 12 month period in patients with inducible sustained ventricular tachycardia and 25% over a 21 +/- 10 month period in patients with inducible ventricular fibrillation. In patients with recurrent unexplained syncope undergoing electrophysiologic testing, a potential cause of syncope is least likely to be found in women without structural heart disease. The results of programmed ventricular stimulation must be interpreted with regard to the method of induction of ventricular tachycardia and the type of ventricular tachycardia induced. The excellent response rate in patients with inducible sustained ventricular tachycardia whose therapy is guided by the results of electropharmacologic testing suggests that sustained ventricular tachycardia is a clinically significant response.(ABSTRACT TRUNCATED AT 250 WORDS)