Current efforts to control tuberculosis include effective treatment as well as early detection. Williams et al. assessed the potential of face-mask sampling as a diagnostic method. They collected face-mask samples from patients with confirmed pulmonary tuberculosis over 24 hours and identified exhaled Mycobacterium tuberculosis using PCR. M. tuberculosis was detected at least four times more frequently in face-mask samples than in sputum samples. Exhaled M. tuberculosis output showed no diurnal pattern, with mostly consistent levels, although high-variable and low-variable outputs were also detected. In a prospective active case-finding pilot study, individuals with symptoms of tuberculosis provided sputum and face-mask samples. Infection was detected in mask samples from individuals who were sputum-negative, but became sputum-positive at 6 weeks. Thus, the approach shows potential for diagnosis and screening.