Abstract. Several methods of heart rate variability have been used to describe different components of the complex regulatory system between heart rate and the autonomic nervous system. Conventionally, heart rate variability has been analyzed with time- and frequency-domain methods. Recently, there has been growing interest in nonlinear methods of heart rate variability as these methods may reveal more delicate changes in heart rate dynamics than the conventional heart rate variability measurements. These nonlinear methods of heart rate variability assess quality properties rather than the magnitude of the signal. Several types of different fractal scaling measures, power-law analyses, complexity measures and symbolic measures have been tested in various patient populations. Methods of analyzing the fractal-like scaling properties and the complexity of the RR interval time series are the nonlinear heart rate variability measures that are most widely studied in clinical settings. In a few recent studies, some of these newer nonlinear measures of heart rate variability have been suggested to be better predictors of mortality than traditional methods. The power-law slope describing long-term fractal-like heart rate behavior has been found to be a better predictor of mortality in the elderly than the conventional measures. More negative values of the power-law slope have been shown to be associated with increased mortality and to precede the spontaneous onset of ventricular fibrillation in postinfarction patients. Altered short-term fractal-like properties of heart rate fluctuations have been observed to have better prognostic power than conventional heart rate variability measures among survivors of acute myocardial infarction with decreased left ventricular function. Furthermore, reduced short-term fractal-like scaling exponent values have been found to be associated with mortality in patients with heart failure. Recent observations in a relatively small group of high-risk patients with implantable cardioverter-defibrillators have shown that the short-term scaling exponent obtained from short-term 10-minute electrocardiographic recordings yields important prognostic information. The short-term scaling exponent has also been suggested to be a powerful predictor of cardiac mortality in the elderly. Reduced approximate entropy values indicating larger predictability and decreased complexity in heart rate behavior have been reported to precede spontaneous episodes of atrial fibrillation in patients after coronary artery bypass surgery and in patients without structural heart disease. Patients with postoperative complications after cardiac surgery have also been observed to have decreased complexity in heart rate dynamics. It has been suggested that the autonomic nervous influences on ventricular repolarization duration and heart rate are qualitively similar in normal individuals, but may be dissimilar in pathological conditions. Repolarization variability has been found to be increased in patients with ischemic and nonischemic dilated and hypertrophic cardiomyopathies as well as in patients with the congenital long QT syndrome. However, data on the nonlinear dynamics of repolarization are very limited. Recent observations suggest that increased temporal complexity of repolarization measured by approximate entropy independently predicts mortality in high risk patients with decreased left ventricular function and implantable cardioverter-defibrillators. None of the temporal repolarization variability parameters determined by traditional statistical measures of variance were associated with adverse events in these patients. Different kind of alterations in nonlinear dynamics of repolarization than in nonlinear dynamics of heart rate are related to risk for mortality in high risk patients with reduced left ventricular function and implantable cardioverter-defibrillators.