Ventricular repolarization (VR) duration is an electrophysiologic parameter that is poorly explored in the conventional electrocardiogram. The heart rate (HR) and the autonomic nervous system (ANS) condition its variations. The static QT measurement in the surface electrocardiogram and its correction using the Bazett formula constitute inadequate tools. The three-dimensional Holter recordings and their computerized analysis are better adapted to assess the dynamicity of VR, a relevant image of the myocardial state. QT and RR values are linearly correlated over 24 hours, but this does not allow one to study specifically the respective roles of HR and ANS on repolarization changes. To do so, it is necessary to select appropriately the QRST complexes according to their environment: not only the last RR cycle but the mean HR over the preceding minutes, and the circadian influences must be controlled to differentiate the role of the short- and long-term influences. Combining the ATREC and the CAVIAR softwares now permits control of both time and space dimensions and to measurement of the QT variations with a precision of the order of 1 ms, a performance necessary to assess tiny but meaningful changes. The QT interval is shorter, and the slope of the QT/RR regression line is steeper at daytime compared with night, and many situations modify this normal behavior. Aging lengthens the QT interval and reduces the day-to-night differences of QT duration and dynamicity. In contrast with duration, dynamicity does not depend on gender. Any heart disease, left ventricular hypertrophy, or heart failure alter QT dynamicity.(ABSTRACT TRUNCATED AT 250 WORDS)