Increased ventricular repolarization dispersion (V RD) has been associated with risk of malignant ventricular arrhythmia. Several ECG parameters have been propose to quantify the V RD, which do not have definitive validation. The objective is to find and validate markers which represent the V RD more accurately. We hypothesized that V RD increment, is reflected by T-wave widening. T-wave duration (T wd) is therefore proposed as a index to quantify V RD. The T wd and the so call T-wave residuum (T wr) have been computed. An isolated rabbit heart model was used. Global V RD, at all myocardium regions, was induced by supplying d-Sotalol and by premature ventricular stimulation (P V S). T wd showed significant increase with V RD increase (78.0±10.3 vs 133.6±29.6 ms with d-Sotalol and 95.2±7.9 vs 118.5±15.7 ms with P V S) while T wr did not change for this global V RD increase.