PURPOSE: To determine the frequency of complications and diagnostic accuracy of CT-guided percutaneous automated gun biopsy, and to compare the results with those reported for fine needle aspiration. MATERIALS AND METHODS: Using automated biopsy devices, 118 CT-guided percutaneous biopsies of pulmonary lesions were performed. An 18-gauge needle was used. Final diagnosis was made with operation or other methods. We retrospectively analyzed the frequency of complications and diagnostic yields of 118 biopsies. RESULT: Four of 118(3.3%) patients developed pneumothorax and two of these required chest tube insertion. Other complications were resolved spontaneously. 106 biopsies (89.8%) yielded sufficient tissue for pathologic evaluation. For cases of malignant and of benign disease, sensitivity was 91.8% and 87.7% respectively ; the corresponding figures for diagnostic accuracy were 88.5% and 78.9%. CONCLUSION: CT-guided automated gun biopsy of the pulmonary lesions is safe, witha pneumothorax rate comparable to that of fine needle aspiration. In the absence of a trained cytologist at the time of biopsy, the diagnostic accuracy of automated gun biopsy of pulmonary lesions compared favorably with the reported accuracy of fine needle aspiration.