A reduced circadian pattern in the QTc interval has been repeatedly reported to provide prognostic information in cardiac patients. However, the results of studies in healthy subjects in which different heart rate correction formulas were used are inconsistent regarding the presence and extent of diurnal variations in QTc. This study compared the diurnal variations in QTc obtained with four frequently used heart rate correction models with those based on individually optimized heart rate correction. In 53 subjects (25 men aged 27 +/- 7 years and 28 women aged 27 +/- 9 years) 12-lead digital ECGs were obtained every 30 seconds during 24 hours. The QT interval was measured automatically by six different algorithms provided by a commercially available device. The QT/RR relation was estimated by four common heart rate correction models and by an individually optimized correction model, QTc = QT/RR alpha. In each 24-hour recording, RR, QT, and WTc intervals of separate ECG samples were averaged over 10-minute intervals. Marked differences were found in the extent of the circadian pattern of QTc obtained with different formulas for heart rate correction. Under and overcorrection of the QT interval resulted in significant over- or underestimation of the circadian pattern. Thus, the extent of circadian variation in QTc depends highly on the heart rate correction formula used. To obtain proper insight regarding diurnal variation in QTc prolongation during pharmacologic therapy and/or to assess higher risk due to impaired autonomic regulation of ventricular repolarization, individualized heart rate correction is necessary.