Inhalation injury is an important contributor to morbidity and mortality in burn victims and can trigger acute lung injury and acute respiratory distress syndrome (ARDS) (1-3). Early diagnosis and treatment of inhalation injury are important, but a major problem in planning treatment and evaluating the prognosis has been the lack of consensus about diagnostic criteria (4). Chest radiographs on admission are often non-specific (5, 6), but indicators include indoor fires, facial burns, bronchoscopic findings of soot in the airways, and detection of carbon monoxide or cyanide in the blood (7). Changes in the lungs may be detected by bronchoscopy with biopsy, xenon imaging, or measurement of pulmonary extracellular fluid (4, 5, 8). These methods have, however, been associated with low sensitivity and specificity, as exemplified by the 50% predictive value in the study of Masanes et al. (8). Computed tomographs (CTs) are better than normal chest radiographs in the detection of other pulmonary lesions such as pulmonary contusion (9, 10). The importance of CT scans in patients with ARDS has been reviewed recently (9), but unfortunately there has been no experience of CT in patients with smoke inhalation injury. To our knowledge, there are only two animal studies reporting that smoke inhalation injury can be detected by CT (4, 11); specific changes in human CT scans have not yet been described. Therefore, confronted with a patient with severe respiratory failure after a burn who from the history and physical examination showed the classic risk factors for inhalation injury, we decided to request a CT.