Prospective identification of patients with structural heart disease who could profit from prophylactic ICD therapy is hampered by the low predictive power of the currently available risk stratification parameters. Microvolt T wave alternans measured noninvasively is a new promising parameter to assess impaired ventricular repolarization which has been associated with an increased incidence of ventricular tachyarrhythmias. T wave alternans is rate-dependent; to induce alternans, heart rate may be increased by atrial stimulation during invasive EP testing or noninvasively by exercise stress testing. The first clinical validation with respect to prediction of inducibility of ventricular tachyarrhythmias and of arrhythmic events during follow-up in patients undergoing invasive EP testing was reported in 1994. Subsequently, a good concordance between the results of invasive and noninvasive assessment of T wave alternans was demonstrated by our group. The first prospective evaluation of the noninvasive alternans measurement using submaximal exercise testing was performed in patients surviving prehospital ventricular fibrillation or sustained ventricular tachycardia referred to our institution. The occurrence of T wave alternans in this patient population was predictive of future tachyarrhythmic events with subsequent appropriate ICD therapy. The results of the currently performed prospective trials in various patient populations will help to establish the utility of T wave alternans assessment as a risk stratifier in clinical practice.