We evaluated the risk, sensitivity, and specificity of transbronchial needle aspiration (TBNA) in a canine model of unilateral Streptococcus pneumoniae pneumonia and compared the results with our previous studies of these parameters in transthoracic needle aspiration (TTN), a catheter-brush (CB), and transbronchial biopsy (TBB) in the same model. The TBNA was performed in 10 dogs with right lower lobe S. pneumoniae pneumonia. The procedure was uncomplicated, but both the sensitivity and specificity were low (3 of 10 cultures yielded S. pneumoniae; only 1 of the 3 was a pure culture). This compares unfavorably with our experience with the other procedures: 9 of 10 cultures were pure with TBB and CB, and the remaining culture was sterile; 10 of 10 were pure with TTN. Given the present unprotected form of the bronchoscopic aspirating needle, TBNA does not appear to be useful in the diagnosis of bacterial pneumonia despite its usefulness in the diagnosis of intrathoracic malignancy.