Active tuberculosis is a serious, sometimes fatal, bacterial infection. Multidrug-resistant strains are associated with higher mortality. Bedaquiline, a diarylquinoline antibiotic, has been authorised in the EU, in combination with other antibiotics, for patients with multidrug-resistant pulmonary tuberculosis. Clinical evaluation of bedaquiline in multidrug-resistant tuberculosis is based on a double-blind, randomised, placebo-controlled trial including 160 patients. During follow-up, there were 10 deaths in the bedaquiline group (including 5 deaths from disease progression) and 2 deaths (both due to tuberculosis) in the placebo group. An analysis of data on 133 patients showed that it took a median of 83 days for sputum cultures to become negative in the bedaquiline group, versus 125 days in the placebo group (p = 0.0004). QT prolongation and hepatic and pancreatic disorders were more frequent with bedaquiline than with placebo. The most frequently reported adverse effects were headache, nausea, arthralgia and pulmonary infections. In practice, bedaquiline reduced the number of contagious patients with multidrug-resistant pulmonary tuberculosis, but mortality was higher than in the placebo group. Its harm-benefit balance is unclear, especially in patients with extensively drug-resistant tuberculosis, for whom there are very few active antibiotics available.