OBJECTIVE: To determine the prevalence and frequency of recurrent ventricular fibrillation (VF) in patients defibrillated by police and firefighters only and to determine its relation to survival. DESIGN: Retrospective observational study. SETTING: Out of hospital. PATIENTS: Individuals with witnessed VF arrest in the Rochester, MN, ambulance public service area who had defibrillatory shocks delivered by police and firefighters and return of spontaneous circulation with shocks only. MEASUREMENTS AND MAIN RESULTS: Electrocardiograms were recovered from data cards in automated external defibrillators used by police (n = 49) or firefighters (n = 18) to deliver shocks from December 1996 through December 2003 in the Rochester, MN, ambulance public service area. Patients with witnessed VF arrest were identified for recurrent VF after initial shock success (first one to three shocks). Both police and firefighters deployed automated external defibrillators delivering nonescalating 150-J biphasic truncated exponential waveform shocks (ForeRunner Automated External Defibrillator, Phillips/Heartstream Operation, Seattle, WA). Among 67 patients, 30 (45%) survived to neurologically intact discharge (overall performance category score of 1 in 29 patients and score of 2 in one patient). Twenty-nine patients (43%) regained spontaneous circulation with shocks only and 25 of 29 (86%) survived. VF recurred in 35 of the 67 patients (52%) while being cared for by police or firefighters. Of these 35 patients, no relation was found between the prevalence or frequency of VF recurrence and survival. CONCLUSIONS: VF recurrence is frequent, variable in time of onset, and unrelated to the performance of bystander CPR. The prevalence and frequency of VF recurrence were unpredictable and do not adversely affect survival. Thus, vigilance for recurrent VF is essential to ensure the survival of patients who are in the care of first responders, even after initial restoration of pulses with shocks.