BACKGROUND: It is generally supposed that diurnal fluctuations of heart rate variability (HRV) are associated with diurnal variability of occurrence of sudden cardiac death, usually caused by tachyarrhythmias. To date, however, there have been no data comparing circadian rhythms of HRV and circadian variability of malignant ventricular tachyarrhythmias in the same patients. MATERIAL/METHODS: We studied 20 consecutive recipients of implantable cardioverter defibrillators in long-term follow-up, comparing the circadian rhythms of HRV in patients with and without typical morning peak occurrence of ventricular tachyarrhythmias. RESULTS: The HRV marker of vagus activity (RMSSD and HF power) displayed physiological circadian variability (highest values during sleep time, morning dip) only in patients without morning peak of arrhythmic events, and inverse circadian variability in patients with tachyarrhythmic morning peak. The variations in LF power and LF/HF ratio in both groups were not statistically significant, but with respect to the time interval between midnight and noon again showed inverse diurnal variation in patients with and without morning peak of arrhythmic events. CONCLUSIONS: These data suggest an association between a reversed circadian rhythm of HRV and morning peak frequency of tachyarrhythmic events in patients with ICD. As pronounced adrenergic hyperactivity in heart failure is the probable reason for paradoxically reversed circadian HRV variation, this association supports the hypothesis that the morning peak frequency of cardiovascular events may be interpreted as a sign of cardiovascular overload by the natural change from sleep to activity after awakening.