Between January 1989 and December 1994 we accomplished 130 CT-guided transthoracic fine needle aspirations (FNA) in 120 patients. Ten patients underwent a second FNA because of the negativity and doubts of the first. Diagnosis was targeted in 114 (95%) patients and 89% showed cytological evidence of malignancy. FNA-CT guided is reliable in the diagnosis of lung cancer, but less accurate in excluding diagnosis of malignancy. In our experience FNA had an accuracy with regard to lung cancer, of 92%. Sensitivity and specificity were respectively 93.8% and 100%. There were 6 false negative and 9 very negative. All patients should have bronchoscopy rather than FNA as the initial diagnostic procedure and perform it only in the absence of endobronchial lesions and malignant cells obtained with cytologic sputum or fiber bronchoscopy. In our experience 91 patients have lung cancer and thoracotomy was performed in 21% with confirm of diagnosis. FNA offers several advantages over other diagnostic procedures used in the evaluation of patients with intra-thoracic nodules and masses. The uses of small needles (20-22 gauge) and CT-guide has practically eliminated the risk of major haemorrhage. Deep and superficial lesions of the lung may be approached safely with FNA-CT guided and complications are no fatalities. Pneumothorax occurred in our experience in 5 cases and no occurred a chest drainage.