Sudden cardiac death (SCD) accounts for ∼50% of mortality after myocardial infarction (MI). Most SCDs result from ventricular tachyarrhythmias, and the tachycardias that precipitate cardiac arrest result from multiple mechanisms. As a result, it is highly unlikely that any single test will identify all patients at risk for SCD. Current guidelines for use of implantable cardioverter-defibrillators (ICDs) to prevent SCD are based primarily on measurement of left ventricular ejection fraction (LVEF). Although reduced LVEF is associated with increased total cardiac mortality after MI, the focus of current guidelines on LVEF omits ∼50% of patients who die suddenly. In addition, there is no evidence of a mechanistic link between reduced LVEF and arrhythmias. Thus, LVEF is neither sensitive nor specific as a tool for post-MI risk stratification. Newer tests to screen for predisposition to ventricular arrhythmias and SCD examine abnormalities of ventricular repolarization, autonomic nervous system function, and electrical heterogeneity. These tests, as well as older methods such as programmed stimulation, the signal-averaged electrocardiogram, and spontaneous ventricular ectopy, do not perform well in patients with LVEF ≤30%. Recent observational studies suggest, however, that these tests may have greater utility in patients with LVEF >30%. Because SCD results from multiple mechanisms, it is likely that combinations of risk factors will prove more precise for risk stratification. Prospective trials that evaluate the performance of risk stratification schema to determine ICD use are necessary for cost-effective reduction of the incidence of SCD after MI.