The vasovagal nature of syncope, which remained unexplained despite full clinical and electrophysiologic investigation, was evaluated by means of 60 degrees head-up tilt test for 60 minutes. Thirty patients (17 men and 13 women, mean age 65 years, 19 with and 11 without organic heart disease) with 1 to 28 (mean 5) episodes of syncope of unknown origin were studied. Head-up tilt test was considered positive if syncope developed in association with hypotension, bradycardia, or both. During baseline head-up tilt 15 patients (50%) had a positive response. Ten patients had a vasodepressor response (marked hypotension without marked bradycardia) and 5 had a mixed response (marked hypotension with marked bradycardia). None of 8 control subjects became symptomatic during the test. Baseline head-up tilt test was positively reproducible in 10 of 14 patients (71%). Nine of these 10 patients underwent serial head-up tilt tests after drug administration to determine the pathogenesis of vasovagal syncope. Atropine prevented tilt-induced syncope in 3 of 8 patients (37.5%), propranolol in 2 of 8 (25%) and etilephrine in 7 of 7 (100%). Seven patients received long-term drug treatment with drugs selected on the basis of acute drug testing. One responder to atropine received transdermal scopolamine and 6 received etilephrine. None of these 7 patients had syncopal recurrences or death during a mean follow-up of 12 months. Head-up tilt is a very sensitive and highly specific test to unmask susceptibility to vasovagal reaction in patients with syncope of unknown origin. Withdrawal of alpha-sympathetic stimulation is a principal mechanism responsible for vasodilation and syncope during head-up tilt.(ABSTRACT TRUNCATED AT 250 WORDS)