Introduction Current resuscitation guidelines were designed to reduce “no flow” time during a resuscitation attempt. However, shocks delivered within 30 seconds of fibrillation onset show higher rates of conversion to pulsatile rhythms (PR); therefore, shortened chest compression durations may benefit refibrillation (refib) cases. Continuous ECG Rhythm analysis during resuscitation attempts, including during CPR, could guide rescuers to deliver early shocks when appropriate, while continuing CPR otherwise. Hypothesis We assessed if a standard AED rhythm analysis algorithm combined with a measurement of myocardial “vigor” could detect shockable rhythms that convert to PR, regardless of ongoing CPR. Methods All delivered shocks (n=954) in a previously described annotated database of records from 165 patients (three EMS systems and two hospitals in five countries) were examined for cases that generated PR (n=64, #pts=27) and for cases of refib (n=323, #pts=87). The period between shocks was analyzed with an AED shock/no-shock algorithm combined with a score related to VF slope and amplitude in a single measure of VF vigor. For refib, detection was considered true positive for rhythms that converted to PR after the next shock if identified 15 seconds prior through 30 seconds after the annotation of refib onset, and false positive if more than 15 seconds preceding onset. Results Of the 64 shocks that converted to PR, 81% (n=51) followed refib that started after the preceding shock, while only 19% (n=12) were fibrillation cases refractory to the previous shock. Excluding two cases with insufficient data to analyze, 48 refib cases that generated PR, after a subsequent shock, were detected within 30 seconds of onset (98% sensitivity). Eleven patients showed false positive detections (87% specificity). Conclusions Automated ECG rhythm analysis can identify onset of refibrillation with high likelihood of PR while avoiding stoppage of CPR for unresponsive rhythms.