PURPOSE: To compare the effectiveness of embolization of the bronchial artery embolization for the management of hemoptysis in pulmonary tuberculosis cases with the severity of lung parenchymal injury and pleural infiltration, as seen on plain chest radiographs, and with the findings of angiography of the bronchial artery. MATERIALS AND METHODS: Among 265 patients with hemoptysis due to pulmonary tuberculosis, the findings of plain chest radiography and angiography of the bronchial artery were comparatively analyzed in the 206 for whom the results of follow up were available. The chest radiographic findings were lassified as follows: Type I refers to simple pulmonary tuberculosis; Type II includes cases in which pulmonary tuberculosis is complicated by bronchiectasis, aspergillosis, or cavitation; Type III is either Types I or II accompanied by pleural infiltrates limited to the lung apex, and Type IV includes cases in which pleural infiltrates have extended beyond the apex in the whole of the lung. Bronchial angiographic findings were divided into four groups : Group I consists of cases which show abnormalities of only the bronchial artery; Group II includes those in which abnormalities are seen in the bronchial artery and either the internal mammary or an ntercostal artery; Group III comprises cases which belong to Group I or II and in which a branch of the subclavian artery is abnormal, and Group IV includes those in which abnormalities occur in at least two branches of the subclavian artery, or there is direct visualization of hypervascularity of this vessel. The initial post-embolic hemostatic effect and the results of follow up were studied over a six-month period. RESULTS: As compared with simple pulmonary tuberculosis (Type I), we found that as the severity of pleural infiltration and complications revealed by plain chest radiographs increased (Type II, III, IV), so did the severity of the manifestation of systemic collateral arteries other than the bronchial artery, as depicted by increase on bronchial angiography. Early post-embolic hemostasis occurred in 96% of Type-I cases (47/49), 82% of Type II (36/44), 70% of Type III (28/40), and 55% of Type IV (40/73). The average success rate was 74% (151/205). During the six month follow-up period, continued hemostasis was found in 80% of Type-I patients (36/45), 75% of Type II (30/40), 59% of Type III (20/34), and 48% of Type IV (20/42). The average long-term hemostasis rate was 66% (106/161). CONCLUSION: Bronchial angiography shows that in systemic collateral arteries circulation increases very substantially, and in cases in which plain chest radiographs depict extensive pleural infiltration or complications associated with pulmonary tuberculosis, it is therefore difficult to expect good hemostatic results after embolization. In such instances we thus recommend aggressive treatment such as surgical intervention.