One hundred eighty-seven cases of cardiac arrest of presumed cardiac etiology were analyzed to determine factors associated with successful out-of-hospital management by paramedic teams. Field and in-hospital records were reviewed to determine the response time of the advanced life support team, the ECG rhythm on arrival, the presence of paramedics on scene at the time of the arrest, whether bystander CPR had been initiated, and the eventual outcome of the resuscitation attempt. A significant difference in survival-to-leave-hospital was seen in patients in whom ventricular fibrillation or ventricular tachycardia (VF/VT) was present on arrival (15.3%) compared to patients with asystole, idioventricular rhythms, blocks, or electromechanical dissociation (3.4%). Survival rates in patients in whom CPR was being performed by a bystander were 24% in the VF/VT group and zero in the "OTHER" rhythms group. When the advanced life support team arrived in less than four minutes, survival rates in the VF/VT group and "OTHER" rhythms group were 23.1% and 7.7%, respectively. When the field team arrived in less than four minutes and a bystander was performing CPR, the survival rates were 42.9% in the VF/VT group and 15.8% in the "OTHER." These data suggest that efforts to improve survival from out-of-hospital cardiac arrest in a community should be directed toward public education, reduction in response times of paramedic units, and lay CPR training.