Cardiac parasympathetic activity was assessed using 24 hour electrocardiographic recordings by measuring the incidence of larger changes in successive RR intervals, which in normal subjects occur frequently but irregularly. In 25 normal subjects the mean number of times per hour in which the change in successive RR interval was greater than 50 ms was 150-250 during waking and 350-450 during sleeping. By contrast, 30 diabetics with medically denervated hearts (12 with cardiovascular reflex evidence of parasympathetic damage and 18 with additional sympathetic damage) and six cardiac transplant patients with surgically denervated hearts had extremely low counts. Additionally, of 20 diabetics with normal cardiovascular reflexes, about half had abnormally low counts, suggesting that this method is better than currently available reflex tests in detecting early cardiac parasympathetic damage. This technique provides a valid and sensitive way of monitoring cardiac parasympathetic activity over prolonged periods.