Objective: We tested the hypothesis that the frequency of shock success differs between initial and recurrent episodes of ventricular fibrillation (VF). Methods: Out-of-hospital cardiac arrest patients presenting with VF from December 1996 to February 2007 defibrillated using an AED with a fixed-energy protocol (150 J) (Philips Medical Systems, Seattle, WA) were included for analysis. We defined shock success as termination of VF within 5 s post-shock (either asystole or organized rhythm). Generalized estimating equation (GEE) analysis was used to adjust for the interrelatedness of shocks within each patient. Results: One hundred and three events occurred during the study period. Patient characteristics included: mean age 64.4 years, 82.5% male, and 81.6% bystander witnessed. Synchronized call-to-shock time was 6.4 ± 2.3 min (mean ± S.D.). VF recurred in 64 (62.1%) patients. Two hundred and fifty-seven shocks delivered for initial (101) or recurrent (156) VF were available for analysis. Initial shocks terminated VF in 93/101 (92.1%); subsequent shocks terminated recurrent VF in 140/156 (89.7%). GEE odds ratio for shock type (initial versus refibrillation) was 1.10 (95% CI 0.37-3.24, p = 0.87). After adjusting for potential confounders, shock type remained insignificant (OR 1.14, 95% CI 0.41-3.2, p = 0.80). We observed no significant difference in ROSC (34.4% versus 46.2%, p = 0.23) or survival (37.5% versus 41.0%, p = 0.72) between those with and without VF recurrence. Conclusions: We observed no significant difference in the frequency of shock success between initial and recurrent episodes of VF using this AED with a 150 J fixed-energy protocol. VF recurrence is common and does not adversely affect shock success, ROSC or survival. © 2008 Elsevier Ireland Ltd. All rights reserved.