The purpose of this study was to evaluate the performance and complication rate of CT-guided percutaneous lung biopsy of pulmonary lesions using a coaxial automated biopsy device and to determine factors influencing the false negative rate for the diagnosis of malignancy. A prospective study of 605 consecutive biopsies performed in 595 patients was undertaken. The performances for the diagnosis of malignancy, for the specific diagnosis of benignity and the complication rate were calculated. Patients' characteristics and lesion and procedure variables were included in a logistic regression model and analyzed as potential predictive variables for occurrence of a false-negative diagnosis of malignancy. The sensitivity, specificity, positive predictive value and negative predictive value for a diagnosis of malignancy were 92.2, 99.1, 99.8 and 73.3%, respectively. A specific diagnosis of benignity was obtained in 43 cases (40%) biopsies. The pneumothorax rate was 17.4% (n=105) with 3 (0.5%) requiring drainage. At multivariate analysis, the sole variable significantly associated with a higher rate of false negative diagnosis of malignancy was lesion size equal to or smaller than 10 mm in diameter. Using a coaxial automated device, CT-guided percutaneous biopsy of pulmonary lesions of size equal to or less than 10 mm in diameter provides a high diagnostic yield with an acceptable complication rate.