All cases of penetrating cardiac trauma in 1985 and 1986 in Jefferson County, Alabama, where patients dying of penetrating trauma received autopsies, were retrospectively reviewed. All hospitals in the county plus the single coroner's office provided the records of the 72 patients comprising this study. Incidents occurred most often in the home or residence (70%) by a known assailant (83%) due to domestic/social disputes (73%). Frequency was greatest in the evening hours (73% between 6:00 PM and 3:00 AM), on weekends in spring and summer. Victims tended to be male (86%), black (72%), married (46%), blue collar workers (62%). There were 41 (57%) gunshot wounds, 3 (4%) shotgun wounds, and 28 (39%) stab wounds with an associated mortality rate of 97%, 100%, and 68%, respectively. Prehospital mortality rate (dead at the scene) was 54.2% (39/72), and death on arrival was 26.4% (19/72), for a combined pretreatment mortality rate of 80.6%. All patients who arrived with no vital signs died. Mortality appeared to be related to mechanism of injury, age, race, sex, vital signs on arrival, number and specific cardiac chambers injured, associated major vascular injury, hematocrit, and mode of transportation. Mortality was not related to caliber of weapon, ethanol level, transport time, time from arrival to operation, or transfusion requirements. There were only ten survivors (1 gunshot wound and 9 stab wounds), all of whom had ventricular injuries and no associated major vascular injuries. The ten survivors represented a 71.4% (10/14) salvage rate for those victims arriving with vital signs. Complications occurred in three patients. Hospitalization averaged 7.3 days in the survivors. Penetrating cardiac trauma remains a serious, socially linked disease with a high rate of mortality. Rapid transport, aggressive resuscitation and cardiorrhaphy remain the best treatment.