To evaluate the diagnostic merit of fiberoptic bronchoscopy in pleural effusions, we performed fiberoptic bronchoscopy in addition to thoracocentesis and closed pleural biopsy in 140 patients who were admitted for diagnostic investigation of the causes of pleural effusions. The patients were divided into subgroups based on clinical features and roentgenographic findings of chest x-ray films. In 39 patients, the pleural effusions were due to various nonneoplastic disorders and in 95 patients it was caused by malignancy. In six patients, the causes of the pleural effusions remained undetermined. A final diagnosis was made by pleural examination in 68 patients, by fiberoptic bronchoscopy in 58 patients, and by either one or both in 100 patients. In 82 patients who had no hemoptysis, a final diagnosis was made by pleural examination in 57 cases and by fiberoptic bronchoscopy in 11 cases only. The diagnostic yield of fiberoptic bronchoscopy (47/58) was superior to that of pleural examination (11/58) in 58 patients presenting with hemoptysis. In 74 patients who had pleural effusions as the sole roentgenographic abnormality, the final entity was established by pleural examination in 45 and by fiberoptic bronchoscopy in 12. The diagnostic merit of fiberoptic bronchoscopy was significantly higher in 59 patients who had concurrent pulmonary abnormalities on their chest roentgenograms. A final diagnosis was made in 43 cases by fiberoptic bronchoscopy in comparison with 21 cases by pleural examination. For patients with unknown pleural effusions, fiberoptic bronchoscopy was more likely to yield a diagnosis than thoracocentesis with closed pleural biopsy in those who had hemoptysis or pulmonary abnormality on chest x-ray films, whereas the reverse applied when these features were absent.