Objective: to evaluate the factors that could effect the risk of pneumothorax in patients undergoing transthoracic biopsy. Material and methods: variables that could increase the risk of pneumothorax were evaluated in 453 CT-guided transthoracic biopsies. Factors were evaluated in two groups: (1) lesion related (presence of emphysema around the lesion, lesion depth, cavitation, presence of fissure/atelectasis and pleural tag in the needle trajectory); and (2) procedure related (biopsy type, needle size, number of passages, level of experience of the operator). All variables were analysed by chi(2) test and multivariate logistic regression statistics. Results: pneumothorax was developed in 85 (18.8%) out of 453 procedures. A chest tube was inserted in ten (11.7%) of them. Variables that were significantly associated with an increased risk of pneumothorax were depth of the lesion (P < 0.001) and severity of the emphysema (P < 0.01). Conclusion: the length of the lung parenchyma traversed during the biopsy is the predominant risk factor for pneumothorax in patients undergoing CT-guided transthoracic biopsy. The risk of pneumothorax was also increased with the severity of the emphysema around the lesion. (C) 2003 Elsevier Science Ireland Ltd. All rights reserved.