Antiarrhythmic drugs prescribed to treat atrial fibrillation can occasionally precipitate ventricular tachyarrhythmia through a prominent slowing of the phase 3 repolarization. The latter results in the triangular shape of ventricular action potential, indicating high arrhythmic risk. However, clinically, the utility of triangulation assessments for predicting arrhythmia is limited owing to the invasive nature of the ventricular action potential recordings. This study examined whether the triangulation effect can be detected indirectly from electrocardiogram (ECG) analysis. Epicardial monophasic action potentials and the ECG were simultaneously recorded in perfused guinea-pig hearts. With antiarrhythmics (dofetilide, quinidine, procainamide and flecainide), a prolongation of the initial repolarization seen in the action potential recordings was closely approximated by lengthening of the interval between the J point and the start of the T wave (the JT start interval) on the ECG, whereas a prolongation of the late repolarization was paralleled by widening of the T wave upslope. Dofetilide, quinidine and procainamide induced a prominent slowing of the phase 3 repolarization in epicardium, leading to triangulation of the action potential. These effects were accompanied by a greater prolongation of the T wave upslope compared to the JT start interval. Flecainide elicited a proportional prolongation of the initial and the late ventricular repolarization, and therefore failed to induce triangulation, based on analysis of both epicardial action potential and ECG profiles.Collectively, these findings suggest that the ratio between the durations of the T wave upslope and the JT start interval may represent the ECG metric of the ventricular action potential triangulation induced by antiarrhythmic drugs.