Aims: Myocardial slices are widely used for cardiac electrophysiology research. However, consistency of electrophysiological properties of the cardiac slices to that of the whole LV has not been studied. The aim of this study is to investigate the differences in depolarization and repolarization between the left ventricle (LV) and the longitudinal ventricular slice using 2D and 3D mathematical models. Methods: Diffusion tensor magnetic resonance imaging data of the canine heart was used to reconstruct LV geometry and fiber orientations. The Luo Rudy I model has been utilized for action potential (AP) generation; the bidomain model has been used for simulation of AP propagation in the tissue and then unipolar pseudo-ECG was calculated. The time of activation and repolarization, repolarization dispersion and dispersion of AP duration were compared between the 2D model of the longitudinal LV slice, 3D model of the whole LV and the LV slice processed from the 3D model. Results: We have obtained the general similarity of depolarization and repolarization between the LV slice and the whole LV. The activation time was 17 ms in the slice, 19 ms in the whole LV and the repolarization time was 331 ms in the slice, 333 ms in the whole LV. We have found local differences in AP parameters and activation time reflecting on the pseudo-ECG between the cardiac slice and that processed from the 3D model. Conclusion: We have shown that the electrophysiological processes in the ventricle and the thin longitudinal ventricular slice are similar, so we believe that cardiac slices can be used to evaluate global electrophysiological properties of the intact heart. The local differences in AP and ECG parameters obtained can be explained by differences in geometry and fiber orientation affecting locally depolarization and repolarization in the myocardium.