To evaluate the utility of sputum induction in the large-scale tuberculosis control program. Prospective study on sputum induction for improving the diagnostic yield of pulmonary tuberculosis, and estimation of the direct costs for sputum induction. Of 1,648 tuberculosis suspects with poor or absent sputum production, induced sputum was smear-positive in 558 patients (353 previously smear-negative, 97 inadequate sputum and 108 unproductive). The direct cost per induced sputum was US $0.37. Sputum induction is an effective, low-cost, and simple technique for improving the smear-positive case detection rate in a tuberculosis control program.