Defibrillation is essential for resuscitating patients with ventricular fibrillation (VF), but shocks often fail to defibrillate. We hypothesized that small variations in pad placement affect shock success, and that defibrillation waveform and shock dose could compensate for suboptimal pad placement. In 10 swine experiments, electrode pads were attached at 3 adjacent anterolateral positions, less than 3 centimeters apart. At each position, 24 episodes of VF were induced and shocked, 8 episodes for each of 3 defibrillation therapies. This resulted in 9 tested combinations of pad position and defibrillation therapy, with 80 episodes of VF for each combination. An episode consisted of 15 seconds of untreated VF, followed by a first shock and, if necessary, a repeat shock. Episodes were separated by four minutes of recovery. Both electrode pad position and therapy order were randomized by experiment. Primary outcome was defined as successful VF termination after the first shock; secondary outcome was the cumulative success of the first and second shocks. First shock efficacy varied widely across the 9 tested combinations of pad position and defibrillation therapy, ranging from 11.3% to 86.3%. When grouped by therapy, first shock efficacy varied significantly between the 3 pad positions: 38.3%, 48.3%, 36.7% (p = 0.02, ANOVA), and, when grouped by pad position, it varied significantly between therapies: 15.0%, 32.5%, 75.8% (p < 0.001, ANOVA). Cumulative 2-shock success varied significantly with therapy (p < 0.001, ANOVA) but not with pad position (p = 0.30, ANOVA). The lowest first shock success was at one position in 6 of 10 animals, at another position in 4 of 10 animals, and never at the third position. Small variations in pad placement can significantly affect defibrillation shock efficacy. However, anatomical variation between individuals and the challenging conditions of real-world resuscitations make optimal pad placement impractical. Suboptimal pad placement can be overcome with defibrillation waveform and shock dose.