BACKGROUND--It has been shown that heart rate variability is decreased in patients with congestive heart failure and that depressed heart rate variability is associated with a propensity to ventricular arrhythmias. Little is known, however, about heart rate variability in patients with both congestive heart failure and ventricular arrhythmias. METHODS--Spectral heart rate variability was analysed from 24 hour ambulatory electrocardiograms in 15 controls, 15 patients with non-sustained ventricular tachycardia associated with clinically normal hearts (NHVT group), and 40 patients with congestive heart failure (CHF group) secondary to either ischaemic heart disease (n = 15) or idiopathic dilated cardiomyopathy (n = 25). Of the 40 patients with congestive heart failure 15 had no appreciable ventricular arrhythmias (ventricular extrasystoles < 10 beats/h and no salvos) and formed the CHF-VA- group. Another 15 patients with congestive heart failure and non-sustained ventricular tachycardia formed the CHF-NSVT group. RESULTS--Heart rate variability was significantly lower in the CHF group than in controls (mean (SD) total frequency 23 (12) v 43 (13) ms; low frequency 12 (8) v 28 (9) ms; high frequency 8 (5) v 14 (7) ms; p < 0.001). The differences in heart rate variability between controls and the NHVT group, between ischaemic heart disease and dilated cardiomyopathy, and between the CHF-VA- and CHF-NSVT groups were not significant. In the CHF group heart rate variability was significantly related to left ventricular ejection fraction but not associated with ventricular arrhythmias. The frequency of ventricular extrasystoles was significantly related to the high frequency component of heart rate variability (r = 0.54, p < 0.05) in the NHVT group. Stepwise multiple regression analysis showed that in the CHF group, heart rate variability was predominantly related to left ventricular ejection fraction (p < 0.05). There was no significant difference in heart rate variability between survivors (n = 34) and those who died suddenly (n = 6) at one year of follow up in the CHF group. CONCLUSION--In patients with congestive heart failure, heart rate variability is significantly decreased. The depressed heart rate variability is principally related to the degree of left ventricular impairment and is independent of aetiology and the presence of ventricular arrhythmias. The data suggest that analysis of heart rate variability does not help the identification of patients with congestive heart failure at increased risk of sudden death.