The limited prognostic value of QT dispersion has been demonstrated in recent studies. However, longitudinal data on physiological variations of QT interval and the influence of aging and sex are few. This analysis included 172 healthy subjects (89 women, 83 men; mean age 38.7 +/- 15 years). Beat-to-beat QT interval duration (QT, QTapex [QTa], Tend [Te]), variability (QTSD, QTaSD), and the mean R-R interval were determined from 24-hour ambulatory electrocardiograms after exclusion of artifacts and premature beats. All volunteers were fully active, awoke at approximately 7:00 AM, and had 6-8 hours of sleep. QT and R-R intervals revealed a characteristic day-night-pattern. Diurnal profiles of QT interval variability exhibited a significant increase in the morning hours (6-9 AM; P < 0.01) and a consecutive decline to baseline levels. In female subjects the R-R and Tend intervals were significantly lower at day- and nighttime. Aging was associated with an increase of QT interval mainly at daytime and a significant shift of the T wave apex towards the end of the T wave. The circadian profile of ventricular repolarization is strongly related to the mean R-R interval, however, there are significant alterations mainly at daytime with normal aging. Furthermore, the diurnal course of the QT interval variability strongly suggests that it is related to cardiac sympathetic activity and to the reported diurnal pattern of malignant ventricular arrhythmias.