Vasovagally mediated cardiac asystole has been proposed as a potential cause of sudden cardiac death. To assess this possibility and identify characteristics that define patients with vasovagally mediated asystole, head-upright tilt-table testing was performed in 50 consecutive patients (26 women and 24 men, mean age 42 +/- 10 years) with recurrent unexplained syncope. The upright tilt-table test was performed in the fasting state for 30 minutes, with or without the use of intravenous isoproterenol (1 to 3 micrograms/min). The production of ventricular asystole lasting greater than 4 seconds was considered a positive result. All patients with tilt-induced asystole received therapy with either beta blockers, disopyramide, transdermal scopolamine or atrioventricular permanent pacing, the efficacy of which was evaluated with serial tilt-table tests. Reproducible tilt-induced asystole occurred in 10 patients (7 men and 3 women, mean age 23 +/- 12 years) (7 patients during baseline tilt, and 3 during isoproterenol infusion). Analysis of this group revealed that they had significantly more frequent and severe syncopal episodes (3 patients had episodes needing bystander cardiopulmonary resuscitation) than did those patients with tilt-induced syncope without asystole. All patients who had tilt-induced asystole eventually became tilt-table negative with therapy (4 with beta blockers, 2 with disopyramide, and 4 with atrioventricular permanent pacing), and over a mean follow-up of 21 +/- 6 months no further syncopal episodes occurred. It is concluded that patients with recurrent tilt-induced asystole represent a distinct subgroup that has recurrent severe syncope that may mimic or result in sudden cardiac death. Thus, the predischarge electrophysiologic study could predict late outcome with recurrence of preexicitation or supraventricular tachycardia in patients who had undergone surgical ablation of the accessory pathway with an overall predictive accuracy of 95% (107 of 113 patients), negative predictive value of 96% (103 of 107), and positive predictive value of 67% (4 of 6).