September, 1992, through May, 1994, thirty patients with hemoptysis were examined with CT, HRCT and bronchoscopy. Our study was aimed at comparing CT and HRCT with fiberoptic bronchoscopy in the identification and assessment of hemoptysis causes and of lesion shape and extent. These data are of basic importance for the interventional radiologist when an intravascular treatment is scheduled. The causes of hemoptysis included cystic fibrosis in 14 patients, bronchiectasis and bronchiolectasis in 11, tuberculosis in 3 and aspergillosis in one. In only one patient the etiology of hemoptysis remained undetected. Among the most common patterns, the "ground-glass" one was the main finding (50%), while bronchiectasis and bronchiolectasis were demonstrated in 40% of the patients. In the extent 10% of cases the cause of hemoptysis was identified with small lesions as a result of previous tubercular infections. Among the causes of hemoptysis, our study included only inflammatory, and not neoplastic, diseases. In 97% of patients, CT and HRCT allowed the diagnosis of lesion type, extent and site, while bronchoscopy did the same in only 35% of patients, because of its lack of accuracy in identifying and characterizing peripheral lesions. Our results suggest that CT and HRCT should be performed after bronchoscopy and before bronchial embolization. Confirming literature data, our study proves CT and HRCT to play a basic role in the diagnosis of the inflammatory conditions causing hemoptysis.