QT dispersion (QTd) has not proven to be a useful marker derived from the 12-lead electrocardiogram (ECG) for stratification of patients at risk for sudden cardiac death. To overcome its methodological shortcomings, novel ECG variables of T wave morphology have been proposed. The total cosine R-to-T (TCRT), T wave morphology dispersion, T wave loop dispersion, normalized T wave loop area, as well as absolute and relative T wave residuum evaluating non-dipolar ECG signal contents were evaluated in two clinical studies involving post myocardial infarction (MI) patients and US veterans with cardiovascular disease. In 280 post MI patients with 27 events over a mean follow-up of 32 months, TCRT and T wave loop dispersion were independent predictors of mortality. In 813 male US veterans with cardiovascular disease the absolute and relative T wave residua were independent predictors of patient risk during a long-term follow-up of more than 10 years. On Cox regression analysis, age, presence of left ventricular hypertrophy (LVH) and left ventricular ejection fraction (LVEF) were also predictors of survival. The latter study in US veterans therefore was the first to demonstrate that a novel parameter characterizing heterogeneity of ventricular repolarization within the 12-lead surface ECG permits risk stratification in patients with cardiovascular disease. All of the ECG variables are easily accessible from digital 12-lead surface ECG recordings using custom computer programs. They may prove useful to identify risk patients that benefit from the implantable cardioverter-defibrillator (ICD).