This simulation study seeks to extend the virtual electrode polarization (VEP) theory for defibrillation to explain the success and failure of biphasic shocks. The goals of the study are to (1) provide insight into why optimal biphasic shocks have a lower voltage defibrillation threshold than monophasic shocks, (2) examine the mechanisms of biphasic shock failure and to determine whether they differ from those of monophasic shocks, and (3) study how the timing of biphasic shock delivery to a spiral wave affects voltage defibrillation threshold. A spiral wave is initiated in a bidomain representation of a 2-cm x 2-cm sheet of ventricular myocardium. The model incorporates nonuniform fiber curvature, membrane kinetics suitable for high-strength shocks, and electroporation. A spatially uniform extracellular field is delivered by line electrodes. The shock establishes VEP that dictates the postshock activity in the tissue. Our results demonstrate that the second phase of biphasic shocks leaves the tissue with substantially smaller postshock excitable gap, thus eliminating the majority of the substrate for reinitiation of reentrant activity. Further, the occurrence of break excitations for weaker biphasic shocks indicates that the mechanisms for biphasic shock failure are more complex than for monophasic shocks. Biphasic voltage defibrillation thresholds range from 8 to 16 V/cm, depending on the position of the spiral wave. An increase in the amount of preshock excitable gap leads to an increase in voltage defibrillation threshold. This study demonstrates the importance of VEP and its interaction with preshock activity in the success and failure of biphasic defibrillation shocks.