To evaluate the airborne viable bacterial concentrations generated during sputum induction and their reduction with exhaust ventilation, ultraviolet germicidal irradiation (UVGI), or both. Exhaust ventilation, upper air UVGI lights, and a portable UVGI unit were operated independently or in combination while and after sputum induction was performed for 58 patients suspected of having active tuberculosis. Viable airborne bacteria were sampled with volumetric air samplers, grown on blood agar, and identified with standard techniques. During and immediately after sputum induction, concentrations of airborne bacteria, particularly respiratory tract or oropharyngeal organisms, increased rapidly, regardless of environmental conditions. The subsequent rate of reduction of airborne bacteria was most rapid with the portable UVGI unit, followed by upper air UVGI with air mixing. Exhaust ventilation achieved high air changes per hour, but efficacy in reducing airborne bacterial concentrations was low. However, the continuous entrainment of bacteria-laden air from the hallway outside may have resulted in underestimation. The efficacy of a wall-mounted upper air UVGI fixture was significantly less if there was no air mixing. The irradiation from this fixture was of adequate germicidal intensity only in a narrow horizontal plane 2.5 m above the floor. Sputum induction was associated with a rapid and substantial increase in airborne bacteria despite the use of exhaust ventilation providing more than 30 air changes per hour, and the adjunct use of UVGI. This emphasizes that healthcare workers involved in similar cough-inducing procedures performed for patients with suspected tuberculosis must wear appropriate personal respirators.