A prospective study was conducted of all cases of attempted resuscitation following cardiac arrest occurring over a period of one year in a tertiary care teaching hospital. Analysis was made of the effects on survival of preselected variables of patients and circumstances of arrest. In 125 cases for which resuscitation was attempted, 49 attempts (39%) were initially successful. Twenty-three patients were discharged from hospital (18% overall survival). The most potent predictors of successful resuscitation were early successful resuscitation (less than 20 mins), age less than 75 years, and cardiac arrest in proximity to an acute ischemic cardiac event (less than 48 h). Survival to discharge was extremely poor when the arrest was unwitnessed (zero of 20 cases) or when the arrest occurred on medical or surgical wards (two of 62 cases). Such low efficacy indicates that reassessment of policy regarding the administration of cardiac resuscitation to patients in hospital is warranted.