Sudden cardiac death (SCD) remains a preeminent public health problem. Identification of high-risk patients, susceptible to SCD, is essential for the successful prophylactic therapy. The majority of such sudden deaths are preceded by fatal ventricular arrhythmias, mainly as a result of ischemic heart diseases. This article is intended to describe the methods of prediction, the results and limitations of the currently used methods. The current measures available for screening of high-risk patients, such as demographic variables, left ventricular contractile function, ventricular ectopy by Holter monitoring, late potentials by signal-averaged ECG, heart rate variability, QT dispersion and even electrophysiologic testing, have limited sensitivity, and specificity, and are only helpful in a minority of patients already at high-risk. The predictive value of each method is modest, even when several predictors are combined. As a result, the effect on the cumulative incidences of SCD, in the population at large, have been relatively small, as the majority of SCD occurs in patients who do not have the characteristics leading to their inclusion in trials of implantable defibrillators. A challenge for the future will be the development of new approaches, or techniques, which will allow screening for markers of increased risk of fatal ventricular arrhythmias in large general populations, of which the relative risk is low, but the number of deaths, due to arrhythmias, are high. Incidences of coronary artery diseases, one of the most important causes of sudden cardiac death, including acute myocardial infarction, have recently grown exponentially in Korea. Therefore, there is a need to develop our own risk stratification strategy by searching for new tools for the prediction, and refinement, of existing tools.