Background Pneumomediastinum after blunt thoracic trauma is often considered a marker of serious aerodigestive injury that leads to invasive testing. However, the efficacy of such testing in otherwise stable children remains unknown. We hypothesize that pneumomediastinum after blunt trauma in clinically stable children is rarely associated with significant underlying injury. Methods We reviewed all patients in our pediatric trauma database (1997-2007) for pneumomediastinum after blunt injury. Patients were then subdivided into 2 groups: group I, isolated thoracic and group II, thoracic and additional injuries. Procedures and imaging were recorded, and outcomes were assessed. Results Thirty-two children with blunt thoracic trauma were included as follows: group I (n = 14) and group II (n = 18). In all patients, there were 28 diagnostic procedures performed resulting in only 1 positive test—a bronchial tear found on bronchoscopy in association with obvious respiratory distress. Group I was more than twice as likely to undergo invasive procedures as group II ( P < .0001), resulting in significantly greater costs (⁎$13683 ± 2520 vs $5378 ± 1000; P < .002). Patients in group I also received more diagnostic imaging to assess pneumomediastinum (1.89 vs 1.08 studies/patient per day; P < .05). More than 28% of all patients were completely asymptomatic and had pneumomediastinum as their only marker of injury. Strikingly, these patients received more than 46% of the procedures. Conclusions Children with pneumomediastinum from blunt trauma often receive invasive and expensive testing with low yield, especially those with isolated thoracic trauma.