We tried to study the relationship between hypotension, clinical findings and hemopneumothorax in penetrating chest injury to determine whether hypotension or clinical findings alone in penetrating chest injury is a reliable indicator for immediate tube thoracostomy. All patients(n=310) with penetrating chest injury admitted to a level I university affiliated trauma center over a two-year period were evaluated. Systolic blood pressure, clinical findings on primary survey and chest x-ray for the presence of hemothorax, pneumothorax or hemopneumothorax were compared. The statistical differences were determined by the Student’s t-test. Of 310 patients, 49 patients died in the trauma bay without chest x-ray, leaving 261 patients. A total of 155 patients (59.4%) had a lesion of hemothorax, peneumothorax, or hemopeumothorax an chest x-ray treated with tube thoracostomy. Hypotension was present in 41 patients (15.75) and hemothorax, pneumothorax, and hemopneumothorax was present in 38 of these 41(92.7%). Hypotension was associated with longer hospital stay, higher mortality an higher Injury Severity Score. Auscultation to detect hemothorax, pneumothorax or hemopneumothorax had a sensitivity of 62.6%, a specificity of 98.1%, an accuracy of 77.0%, and a positive predictive value of 98.0%. Penetrating chest injury with hypotension or decreased breathing sounds alone has a significant incidence of hemothorax, pneumothorax, or hemopneumothorax on the side of the injured hemothorax. We conclude that immediate tube thoracostomy before chest x-ray is warranted.(Tzu Chi Med J 1996; 8: 263-269)