Despite its historical decline, TB remains a significant cause of infectious disease-related global deaths. The lack of reliable diagnostic tests for vulnerable groups, such as children and immunocompromised patients, remains a challenge for TB control. For decades, it has been recognised that exhaled breath has great potential as a non-invasive and universally accessible clinical alternative to sputum and invasive sampling methods. Although translation into clinical practice has not yet occurred, there has been significant progress with promising results in various applications, including diagnosis, estimation of infectiousness, and monitoring of treatment response. More recently, the COVID-19 pandemic reignited global interest in this field and technological advances have further accelerated its development. In the coming decade, breath sampling will enhance our understanding of respiratory infectious diseases and host-immune responses, which may lead to clinical applications. Here we discuss the diagnostic landscape of TB and the current state of the art of breath sampling.