The most frequently proposed mechanism for vasodepressor syncope is based on cardiac mechanoreceptor activation by augmented sympathetic tone. Because of the central role of the autonomic nervous system in this response, we hypothesized that the responses of the sympathetic and parasympathetic nervous systems (as assessed by analysis of heart rate variability) to orthostatic stress would differentiate patients with a positive from those with a negative tilt-table response. We therefore evaluated 28 patients undergoing tilt-table testing for presumed vasodepressor syncope. Based on 5-minute electrocardiographic samples obtained during the supine and upright phases (without isoproterenol infusion), we computed the mean RR interval, reflecting integrated cardiac sympathetic and parasympathetic tone, as well as the root-mean-square of successive differences of the RR intervals (RMSSD), a measure of high-frequency heart rate variability that is correlated with parasympathetic tone. Eleven patients had a negative and 17 a positive tilt response. There were no differences between the groups at baseline. In response to upright tilt, the mean RR decreased by a similar magnitude in both groups. In contrast, RMSSD decreased by 36% (p = 0.05) in response to upright tilt in patients with a negative response, but did not change significantly in patients with a positive tilt response. Absence of a decrease in RMSSD in response to orthostatic stress had 100% specificity and 41% sensitivity for predicting a positive test result. Thus, failure of withdrawal of parasympathetic tone (as assessed by RMSSD) during upright tilt predicts a positive tilt response.(ABSTRACT TRUNCATED AT 250 WORDS)