Many studies of prehospital resuscitation report on selected populations. We examined a series of 445 unselected nontraumatic cardiac arrests. Emergency cardiac care (ECC) was not initiated in 126 (28%). ECC was begun in 319 (78%), but was terminated in 132 (33%). Ninety-four (21%) were admitted to the hospital with palpable pulses and organized rhythm (successful resuscitation/save rate for patients presenting in ventricular fibrillation was 50%/25%. Multivariate regression analysis was used to identify the relative importance of significant variables in predicting survival, and the analysis identified the presence of ventricular fibrillation, short paramedic response times, and short paramedic treatment times.