Eight hundred forty-seven consecutive patients discovered in cardiac arrest by first responding firefighters received initial defibrillation attempts using automatic external defibrillators. The effect of electrode polarity on defibrillation and resuscitation was determined in the subset of 289 (34%) with ventricular fibrillation in a prospective, randomized trial. The ECG was recorded in 205 consecutive patients whose initial rhythm was ventricular fibrillation. Eighty-seven of 114 patients (76%) in whom the apex chest electrode was positive were defibrillated with the first 200-joule shock, compared to 70 of 91 patients (77%) in whom the apex electrode was negative. There was no difference in the type of rhythm established, e.g., organized versus brady-asystole following defibrillation with either electrode polarity. Resuscitation was possible in 56% of patients in whom the apex electrode was positive and 60% of those in whom the apex electrode was of negative polarity. Hospital survival rates (26% vs 27%) were also similar for both treatment groups. Unlike results during experimental external defibrillation of animals or those obtained using implantable defibrillators, this randomized trial of external defibrillation conducted during attempted out-of-hospital resuscitation showed no difference in outcomes related to electrode polarity.