Risk stratification for primary prevention of sudden cardiac death (SCD) remains a major challenge in cardiology. The utility of T-wave alternans (TWA) as a marker of risk of life-threatening ventricular tachycardia and fibrillation is supported by two decades of basic and clinical research. Both frequency- and time-domain methods have been developed, validated, and made available in clinical practice. A principal application of TWA testing has been to improve assessment of patients with depressed left ventricular ejection fraction (EF; </= 40%) who are considered for implantable cardioverter-defibrillator (ICD) implantation for primary prevention of SCD. TWA has been most useful in identifying patients who are unlikely to benefit from ICD therapy. Although patients with low EF should remain an important focus, the absolute number of SCD events is far greater among post-myocardial infarction patients with relatively preserved EF, even though the incidence of SCD in this population is low. Recent studies suggest that TWA testing is predictive in this population as well. Absolute quantification of TWA rather than binary classification into "normal" or "abnormal" appears to be valuable in more finely stratifying the magnitude of arrhythmic risk. Longitudinal testing may be warranted in certain populations, although the optimum interval remains to be determined. Combining TWA with noninvasive markers of autonomic function, such as heart rate turbulence, may further increase predictive accuracy. Future development will likely expand the role of TWA testing with routine exercise and ambulatory electrocardiographic monitoring to screen lower-risk populations and to guide medical and device-based therapy.