A retrospective review was performed of 98 patients admitted to a Level I Trauma Center between July 1989 and December 1990 with a diagnosis of blunt chest trauma. Of these patients, 49% suffered either immediate or delayed complications. Immediate complications included hemothorax or pleural effusion in 21.4%, pneumothorax in 36.7%, and ruptured diaphragm in 2%. Delayed complications occurred in 29.6% of patients overall; these included pulmonary contusion in 33.7%, pneumonia in 5.1%, and pulmonary embolus in 1.0%. Both initial Trauma Score (TS) and the Injury Severity Score (ISS) were predictive of mortality. The mortality rate was significantly greater in patients with a TS < or = 13-30.2% of these patients died of their injuries as opposed to 1.9% of those with a TS > 13. Similarly, 19.7% of patients with an ISS > or = 16 died, while all patients with an ISS < 16 survived. These differences were statistically significant (p < 0.001 and p < 0.05, respectively). Neither TS nor ISS, however, was predictive of either immediate or delayed complications. Even with a TS > 13, 47.2% of patients suffered complications; 80% of these patients had delayed complications. Similarly, 51.8% of patients with ISS < 16 had complications. Statistically, neither TS nor ISS could be used to predict complications, regardless of the reference value chosen for TS or ISS. Therefore, it is imperative that all patients with blunt trauma be considered at risk for secondary complications, even those with "favorable" TS or ISS.