Medical records and computed tomography (CT) images were reviewed retrospectively for 30 animals (27 dogs, two cats, one cougar) in which CT-guided intrathoracic fine-needle aspirates (FNA) (12), core biopsies (10) or both (8) were performed. Sample interpretation was listed as diagnostic or nondiagnostic and nonneoplasia or neoplasia. Diagnostic results were inconclusive in 35% FNA and 17% biopsies. FNA and biopsy interpretations were in agreement in seven patients, one nonneoplasia, and six neoplasia. A clinical diagnosis was made in 65% FNA and 83% biopsies. When 18 patients with confirmed diagnoses were used, overall accuracy for diagnosis was 92% for FNA and biopsy and the sensitivity for neoplasia was 91% using fine needle aspirate and 80% using biopsy. Complications seen on CT images were noted in 43% of patients, four pneumothorax, five pulmonary hemorrhage, and four with both. No clinical manifestations were noted and treatment was not necessary. Significant correlation was noted between complications and penetration of aerated lung, but not with lesion location, type of disease, method of sampling, width of mass and depth of aerated lung penetrated. CT-guided sampling is relatively safe and useful in the diagnosis of intra-thoracic lesions, especially neoplasia. FNA samples are nondiagnostic more often than biopsy samples. Sub-clinical pneumothorax and hemorrhage are common when aerated lung is penetrated.