Attenuated cardiac parasympathetic activity appear to be an important risk factor contributing to sudden cardiac death in subjects with overt coronary disease but its predictive value in otherwise healthy normal subjects is not known. We have for 8 years followed 260 apparently healthy adult subjects who underwent Holter monitoring. Twelve died, 14 developed ischaemic heart disease and four suffered sudden cardiac death. A healthy control subject was matched, along with other risk factors, for each case. In each subject 24-h heart rate variability was calculated as the deviation of all normal R-R intervals from mean R-R (SD) and the percentage of successive R-R interval differences exceeding 6% (%DIF6%)--this was used as an index of cardiac parasympathetic activity. There were no significant differences in heart rate variability between the cases developing problems and controls. In the sudden cardiac death victims, however, there was a clear trend towards lower heart rate variability. In them waketime mean SD was 73 ms versus 85 ms for cases and controls respectively (p = 0.08), and for sleeptime 61 ms versus 76 ms (p = 0.07). Compared to normal limits for heart rate variability obtained in 140 subjects that remained healthy for 8 years, figures for both SD and %DIF6% in sudden cardiac death subjects were at or below 95% confidence limits. The results indicate that altered autonomic balance may contribute to sudden cardiac death even in apparently healthy subjects. Subjects with a low 24-h heart rate variability on Holter monitoring may be predicted at an early stage of being at greater risk.(ABSTRACT TRUNCATED AT 250 WORDS)