The charts of 71 patients admitted to one teaching institution over a 4-year period with a primary or secondary diagnosis of "rule out cardiac contusion" and of another 62 admitted to a second institution with more severe injuries and suspicion of cardiac contusion were reviewed to determine if mortality or morbidity would have occurred if all patients with normal electrocardiograms (EKGs) in the emergency department (ED) were discharged (or admitted to unmonitored beds for other injuries). Only 13 patients developed cardiac problems: two elderly patients died in the ED, while the others experienced arrhythmias or, less commonly, pump failure requiring treatment or observation. All 13 had EKG changes present while still undergoing evaluation in the ED: 11 had a specific problem on arrival, 1 developed a problem while still being evaluated in the ED, and the 13th had what was probably an iatrogenic problem. Importantly, 5 of 12 patients had normal creatine phosphokinase-MB fractions, and 5 of 9 had normal echocardiograms. No patient with a normal EKG had subsequent cardiac problems. Operative intervention for other injuries was necessary in 26 patients overall, and there was no cardiac morbidity. We conclude that had the EKG been used as the sole screening tool, approximately 25% of these patients could have been discharged from the ED without missing problems. In addition, management would have been greatly simplified, and the hospital would have realized substantial savings, both in terms of direct costs and in the freeing of valuable and scarce resources.