Drug-resistant tuberculosis (DR-TB) in adults is either acquired due to poor treatment management or transmitted from infectious DR-TB cases, while children mainly have transmitted disease. Diagnosis of DR-TB relies on drug susceptibility testing (DST), which is not routinely performed in high tuberculosis (TB) burden settings. The Category II retreatment regimen is inadequate for Category I failures if multidrug-resistant TB (MDR-TB) is present. Where possible, DST should be performed for Category I failures and other patients with a high risk of DR-TB. Fluoroquinolones (FQs) should be used with caution in the treatment of mono- and polyresistant TB. Modification of regimens for mono- and polyresistance is prone to error if DST results are not reliable or if there is possible additional drug resistance due to further drug exposure. While standardised treatment is feasible for MDR-TB patients never previously treated with second-line drugs, a different strategy is required for those MDR-TB patients who have previously been treated with second-line drugs. Sputum conversion, adverse effects and adherence to MDR-TB treatment should be monitored closely. The presence of FQ resistance prior to MDR-TB treatment poses a serious challenge. To prevent the development of extensively drug-resistant TB, strategies to protect the FQs, the most important second-line agents, need to be developed. Clinical trials assessing MDR-TB treatment regimens are urgently needed.