The relatively high incidence of endobronchial tuberculosis involvement in children suggests that bronchoscopy is extremely useful in the early diagnosis of tuberculous pneumonia. Pertinent bronchoscopic findings reported are granulation tissue, obstructive caseum, mucosal inflammation, and bronchial stenosis.2 Analysis of bronchoalveolar lavage (BAL) fluid is most useful for the isolation of M tuberculosis in sputum-negative pulmonary tuberculosis of adult patients. However, in a recent study of infected children comparing gastric and BAL cultures, M tuberculosis was isolated from 50% of gastric lavage samples vs 10% of BAL specimens indicating that gastric lavage remains the method of choice for isolation of tubercle bacilli in young children.6 The chemotherapy of tuberculosis is directed toward eradication of tubercle bacilli and prevention of emergence of drug-resistant strains of M tuberculosis, using multiple drugs for 6 to 9 months.7 There is a high likelihood of developing bronchial stenosis after recovery from active disease. Though corticosteroids have been used as an adjunct in treatment of tuberculous meningitis and pleural or pericardial effusions, their role in the treatment of endobronchial tuberculosis remains controversial.2,4 Endobronchial abnormalities have responded favorably to prednisolone therapy in 6 of 8 children treated,2 but Nemir et al8 reported corticosteroids were ineffective in reducing the incidence of residual fibrosis in children. Because endobronchial lesions are more commonly present in patients with prolonged undiagnosed or inadequately treated tuberculosis, optimal management requires early diagnosis and adequate treatment of the primary pulmonary disease.