Over a 20-month period, we prospectively studied 41 children with blunt thoracic trauma (BTT) to determine the incidence and morbidity of cardiac injury in this population. Four patients died, and the data were incomplete in 10, leaving 27 for analysis. Serial electrocardiograms (ECG) were abnormal in 36% of the patients studied, serial creatine phosphokinase isoenzyme (CK-MB) ratios were elevated in 31%, echocardiogram showed septal dysfunction in 14%, and pyrophosphate scan showed grade 1 uptake in 14%. There was poor correlation among tests, since only four patients had more than one abnormal test. There was no significant difference in Injury Severity Score or Thoracic Abbreviated Injury Scale between patients with and without abnormal investigations. No patient in the study developed arrhythmias or cardiac failure. In 13 trauma-related deaths undergoing autopsy over the same period, including the four with thoracic trauma, none had evidence of cardiac injury. These results indicate a striking lack of consistency in the diagnosis of posttraumatic cardiac injury in children using standard investigations. The absence of adverse cardiac events in surviving patients and the lack of autopsy evidence of cardiac injury in the trauma deaths suggest that the actual incidence and clinical significance of these lesions in children is lower than generally reported. Children with BTT should be followed clinically, and reliance on screening tests should be avoided.