ABSTRACT The efficacy of many compounds against Mycobacterium tuberculosis is often limited when administered via conventional oral or injection routes due to suboptimal pharmacokinetic characteristics. Inhalation delivery methods have been investigated to achieve high local therapeutic doses in the lungs. However, previous models, typically employing wild-type M. tuberculosis strains, were intricate, time-consuming, labor-intensive, and with poor repeatability. In this study, we developed an autoluminescence-based inhalation administration model to evaluate drug activity by quantifying relative light units (RLUs) emitted from live mice infected with autoluminescent M. tuberculosis . This novel approach has several improvements: it eliminates the need for anesthesia in mice during administration and simplifies the instrument manipulation; it is cost-effective by utilizing mice instead of larger animals; it shortens time from several months to 16 or 17 days for obtaining result; it is non-invasive by measuring the live RLUs of mice; up to six mice can be administrated daily and simultaneously, even 2-3 times/day; results are relatively objective and repeatable minimizing human factors. Proof-of-concept experiments demonstrated that inhalable rifampicin, isoniazid, and ethambutol showed anti- M. tuberculosis activity at concentrations as low as 0.5, 0.5, and 0.625 mg/mL, respectively, as evidenced by comparing the live RLUs of mice. Furthermore, consistency between RLUs and colony-forming units of the lungs reaffirms the reliability of RLUs as an indicator of drug efficacy, highlighting the potential of this approach for accurately assessing anti- M. tuberculosis activity in vivo . This autoluminescence-based and non-invasive inhalation model offers a substantial reduction in the time, effort, and cost required for evaluating the efficacy of screening new drugs and repurposing old drugs in vivo via inhalation administration.