The goal of this modeling study is to investigate the mechanisms responsible for the upper and lower limits of vulnerability (ULV and LLV) to re-entry induced by electric shocks within the three-dimensional volume of the heart. We use a geometrically accurate rabbit ventricular model with realistic fiber architecture that also includes the blood in the cavities and a perfusing bath. The shocks are delivered over a range of strengths and coupling intervals via two large mesh electrodes located at the vertical boundaries of the perfusing bath. Our results demonstrate that shock-induced virtual electrode polarization (VEP) in the midmyocardium is weaker and more complex than VEP on the surfaces, where only 2 areas, one of positive and one of negative polarization, are induced. Transmural views of the ventricles show that, in all cases, tissue in the LV free wall and in the septum is deexcited by the shock providing an excitable path for wavefront propagation. Conversely, the RV free wall myocardium is depolarized after the end of the shock. The evolution of postshock electrical activity in the RV free wall plays a critical role in determining the outcome of the shock. In all cases, a wavefront starts in the apex at the site of largest transmembrane voltage gradient between oppositely polarized areas. For shocks of strength above the LLV, the postshock refractoriness of the RV free wall produces the unidirectional block necessary for reentry induction. If shock strength is below the ULV, the RV free wall recovers in time to provide the reentrant pathway. In contrast, for shocks of strength above the ULV, the postshock excitable gap in the LV free wall and in the septum is depolarized before the RV free wall recovers. Therefore, both ventricles are refractory and reentry is not induced