Heart rate variability was measured in 77 healthy controls and 343 diabetic patients by a count of the number of beat-to-beat differences greater than 50 ms in the RR interval during a 24 hour ambulatory electrocardiogram. In the healthy controls the lower 95% tolerance limits for total 24 hour RR interval counts were approximately 2000 at age 25, 1000 at 45, and 500 at 65 years. Six controls confined to bed after injury had normal 24 hour patterns of RR counts, while eight other controls showed loss of diurnal variation in both heart rate and RR counts during a period of sleep deprivation. RR counts in ten controls on and off night duty increased during sleep whenever it occurred. Nearly half (146) the 343 diabetic patients had abnormal 24 hour RR counts. The percentage of abnormal RR counts increased with increasing autonomic abnormality assessed by a standard battery of tests of cardiovascular autonomic function. A quarter of those with normal cardiovascular reflex tests had abnormal 24 hour RR counts. There were close correlations between 24 hour RR count results and the individual heart rate tests (r = 0.6). The assessment of cardiac parasympathetic activity by 24 hour RR counts was reliable. The diurnal variations in RR counts seen in the controls were probably related to sleep rather than either posture or time of day. The method was more sensitive than conventional tests of cardiovascular reflexes.