tive laboratory data, respiratory function, and an electrocardiogram were all within normal ranges. After the patient entered the operating room, an epidural catheter was placed at the 8th and 9th thoracic level (T8-9 interspace), and no blood or central spinal fluid was drawn. General anesthesia was induced with fentanyl 0.1 mg, thiopental 250 mg, and vecuronium bromide 9 mg, and the trachea was intubated.Anesthesia was maintained with N 2 O (50%), O 2 (50%), sevoflurane (1%-1.5%), and intermittent epidural administration of 1% lidocaine. The patient was mechanically ventilated with tidal volumes of 450-550 ml (end-tidal CO 2 , 35-40 mmHg) and an inspiratory peak pressure of 14-18 mmHg. A catheter was inserted via the left radial artery for arterial blood pressure monitoring and blood gas analysis. Arterial blood gas analysis before skin incision under 50% oxygen revealed pH 7.42, PaO 2 190 mmHg, PaCO 2 37.6 mmHg, and SaO 2 100%.About 2 h after the start of the surgery, we noticed a decrease in PaO 2 to 88 mmHg (pH, 7.39; PaCO 2 , 39.4 mmHg). Intratracheal suctioning was carried out using a bronchofiberscope, and we drew a considerable quantity of secretion from both lungs. The position of the tracheal tube was correct, and a bronchofiberscopic examination showed no abnormalities in either lung. We then considered the possibility that the pneumothorax was a complication of the central venous catheter insertion, but this possibility seemed unlikely because of the absence of diminished respiratory sounds in the upper and middle lung field. Because the operation was scheduled to finish within another 30 min, we decided to proceed conservatively and gave the patient 100% O 2 .A chest radiograph was taken immediately after the end of surgery, and we observed that the right lower lung field had collapsed as a result of a pneumothorax of the inferior lobe, and that no major collapse or abnormality in the apex and middle lung field had occurred