Purpose : To compare the precise roles of high-resolution computed tomography (HRCT) and fiberopticbronchoscopy (FOB) in the evaluation of patients presenting with hemoptysis and to determine the optimal timingfor HRCT. Materials and Methods : The results of HRCT and FOB were compared in 23 patients (15 men, 8 women)presenting with hemoptysis. Etiologies included bronchietasis(n=4), parenchymal pulmonary tuberculosis(n=4), lungcancer(n=4), endobronchial tuberculosis(n=2), and broncholithiasis(n=2). Hemoptysis was proved to be due tomiscellaneous causes in an additional three cases and to be cryptogenic in four. The diagnostic results of FOBperformed before and after HRCT were compared as were those of HRCT performed within and after the first 48 hoursof active bleeding. Results : FOB and HRCT offered a correct diagnosis in 39% and 65% of cases, respectively(p=0.005). HRCT demonstrated three cases of bronchiectasis and three of parenchymal pulmonarytuberculosis which were beyond the range of a bronchoscope. In two of five cases in which HRCT findings werenonspecific, chondromatous hamartoma and lung cancer were confirmed by FOB. In cases where HRCT was performedprior to FOB, the latter demonstrated the location and diagnosis in 82% and 47% of cases, respectively (p=0.303) ;when HRCT was performed after FOB, HRCT was correct in 67% and 17% of cases, respectively (p=0.178). In none ofthree cases (0%) in which HRCT was performed during the first 48 hours of active bleeding did the procedure allowa specific diagnosis. In 15 of 20 (75%) cases in which HRCT was performed after the first 48 hours, however, thediagnosis provided by CT was correct. Conclusions : The results of this study suggest that in patients presentingwith hemoptysis, both HRCT and FOB should be used for evaluation, since they are diagnostically complementary. FOBis more useful for the diagnosis of endobronchial lesion, and HRCT for bronchiectasis and parenchymal pulmonarytuberculosis. If, in cases of hemoptysis, initial diagnosis is attempted within the first 48 hours of activebleeding, FOB should be the initial step, and HRCT images should not be obtained until active bleeding has beenshown on plain chest radiograph to have abated. If this initial approach takes place after the first 48 hours ofactive bleeding, FOB and HRCT are equally suitable.