Introduction: Unsuccessful defibrillation shocks may reinitiate fibrillation by causing postshock reentry. Methods and Results: To better understand why some waveforms are more efficacious for defibrillation, reentry was induced in six dogs with 1-, 2-, 4-, 8-, and 16-msec monophasic and 1/1- (both phases 1 msec) 2/2-, 4/4-, and 8/8-msec biphasic shocks. Reentry was initiated by 141 ± 15 V shocks delivered from a defibrillator with a 150-μF capacitance during the vulnerable period of paced rhythm (183 ± 12 msec after the last pacing stimulus). The shock potential gradient field was orthogonal to the dispersion of refractoriness. Activation was mapped with 121 electrodes covering 4 × 4 cm of the right ventricular epicardium, and potential gradient and degree of recovery of excitability were estimated at the sites of reentry. Defibrillation thresholds (DFTs) were estimated by an up-down protocol for the same nine waveforms in eight dogs internally and in nine other dogs externally. DFT voltages for the different waveforms were positively correlated with the magnitude of shock potential gradient and negatively correlated with the recovery interval at the site at which reentry was induced by the waveform during paced rhythm for both internal (DFT = 1719 + 64.5∇V - 11.1RI; R2 = 0.93) and external defibrillation (DFT = 3445 + 150∇V - 22RI; R2 = 0.93). Conclusion: The defibrillation waveforms with the lowest DFTs were those that induced reentry at sites of low shock potential gradient, indicating efficacious stimulation of myocardium. Additionally, the site of reentry induced by waveforms with the lowest DFTs was in myocardium that was more highly recovered just before the shock, perhaps because this high degree of recovery seldom occurs during defibrillation due to the rapid activation rate during fibrillation.