In-vitro studies indicate that the electric-field intensity, or voltage gradient (VG), generated by a defibrillation shock is a determinant of defibrillation success as well as potential shock-induced cardiac injury. It is not clear how common descriptors of shock dose, e.g., joules (J), relate to VGs. Objectives: To assess the relationships between shock energy descriptors and VG. Methods: One monophasic and three biphasic waveforms were compared using transthoracic shocks and standard electrodes in five swine. VG measurements via intracavitary multielectrode-tipped catheters were compared with delivered energy and peak current. Shock variables were recorded at native transthoracic impedance and an adjusted impedance approximating that of typical humans. Results: For shocks at the same energy setting, peak current and VGs varied widely among the four defibrillators. At simulated human impedance (75 V), shocks at each device's maximum energy setting produced similar VGs among the three biphasic defibrillators, despite different delivered energies. For both native impedance (35 V) and at 75 V, VG correlated with peak current (r = 0.81 and 0.77, respectively) but not energy setting (r = 0.61, 0.52) or delivered energy (r = 0.58, 0.56). At the maximum energy setting of each device, maximum recorded VGs for both monophasic (33 V/cm) and biphasic (24 V/cm) defibrillators were less than those reported to cause myocardial injury (>60 to 80 V/cm). Conclusions: Energy descriptors correlate poorly to actual shock intensities. When compared with reported VG thresholds of myocardial injury, this study suggests that risk of injury from critically strong VGs is low for all of these defibrillators and equivalent among tested biphasic waveforms.