The effect of 24-hour sleep deprivation on cardiac vagal tone during the first 90 minutes of sleep was studied in 16 subjects using a new real-time index of cardiac parasympathetic activity. The level of cardiac vagal tone in subjects who were sleep deprived was more than doubled that in non-sleep-deprived subjects during sleep stage 0 immediately prior to the onset of sleep stage 1. The mean cardiac vagal tone fell significantly (p < 0.01), from a high level of 19.2 arbitrary units of a linear vagal scale (LVS) to 10 units on reaching slow-wave sleep stage 4 in sleep-deprived epileptic subjects, There was a similar fall in cardiac vagal tone from 18.7 units to 12 units in the LVS in sleep-deprived normal subjects, but this was not statistically significant. There was no change in the mean vagal tone of non-sleep-deprived normal subjects. We propose that the cardiac vagal tone can be used as an objective index of sleep propensity.