PURPOSE: To compare the CT emphysema score with various factors of pulmonary function test by simple spirometry and to use the result as a predictor of pneumothorax in percutaneous transthoracic fine needle aspiration biopsy. MATERIALS AND METHODS: The CT scans of 106 patients who had undergone percutaneous transthoracic fine needle aspiration biopsy of lung lesions within the previous 18 months were retrospectively reviewed. In 75 of these 106 cases, the results of the pulmonary function test were also reviewed. On plain chestradiography, pneumothorax was noted in 20 cases (19%). Emphysema was blindly evaluated. We divided each lung into four segments and determined the severity and involved volume of emphysema, as seen on CT. Sseverity was classified as one of four grades, as follow: absence of emphysema=0; low attenuation area of less than 5mm=1; low attenuation area of more than 5mm, and vascular pruning with normal lung intervening=2; and diffuse low attenuation without intervening normal lung, and larger confluent low attenuation with vascular pruning and distortion of branching pattern occupying all or almost all the involved parenchyma=3. The involved area was also classified as one of four grades: less than 25%=1; 25-49%=2; 51-74%=3; and more than 75%=4. The CT emphysema score was defined as the average of the grade of severity multiplied by the grade of involved area. Pulmonary function tests, consisting of simple spirometry and a pulmonologist's interpretation, were evaluated. We also evaluated depth and size of lesion as known predisposing factors in postbioptic pneumothorax. Statistical analysis was performed using the chi-square test, Wilcoxon ranks sum W test and the student t test. RESULTS: A comparision between the two groups of occurrence (with or without pneumothorax) showed the emphysema scores to be 1.69±2.0 and 1.11±2.9, respectively; there was thus no significnt difference between the two groups (Z=-0.048, p>0.10). Nor were differences revealed by the pulmonary function test; the scores were 2.45±1.2 versus 2.23±0.8 in FEV1, and 73.4±14 versus 78.8±13 in FEV1/FVC. In the group with pneumonthorax, the lesion was significantly deeper, however, (1.92±1.6 cm versus 0.75±1.0cm; Z=-3.50, p<0.01) and the lesion was significantly smaller (3.37±1.7cm versus 4.20±2.0cm; Z=-1.86, p<0.10). In the pneumothorax group, the CT emphysema score was also significantly higher (1.94±3.9 versus 0.39±1.8; Z=-1.513, p<0.10). CONCLUSION: For the prediction of pneumothorax in percutaneous transthoracic fine needle aspiration biopsy, the CT emphysema score is more useful than the pulmonary function test.