The results of emergency room thoracotomy (ERT) and cardiorrhaphy for 91 patients with penetrating cardiac injuries admitted in extremis to Lincoln Medical and Mental Health Center from 1963 to 1981 are reviewed to determine criteria for selection of patients for this procedure. Four groups were defined based on the severity of the effects of their injuries. The survival rates were 32.1 and 33.3%, respectively, for Group I (‘fatal’) and Group II (‘agonal’) patients. There were no survivors in Group IV (‘D.O.A.’) patients for whom ERT is a fruitless procedure. Survival in Group III (‘profound shock’) patients was only 40%, which might have been improved if ERT had been performed without delay. We conclude that ERT is essential for patients with ‘fatal’ and ‘agonal’ wounds and advise prompt ERT for patients in ‘profound shock’ who do not respond immediately to rapid volume infusion.