Development of new drugs to treat tuberculosis (TB) faces even more constraints than the development of therapeutic agents for other diseases. This is due, in part, to intrinsic properties of the tubercle bacillus, such as its slow growth, phenotypic drug resistance during persistence and the need for compounds with a novel mode of action because of the increasing prevalence of primary resistance to the current TB drugs. Demographic changes to the population of TB patients are also a confounding factor; these now include co-infection with HIV, but other elements, such as the growing type-2 diabetes epidemic, should not be ignored. Consequently, a new TB drug will not only have to pass all the safety requirements associated with prolonged administration but also have to be compatible with antiretroviral therapy and, possibly, other medications. Here, we review the changing clinical landscape of TB and outline how this needs to be taken into consideration when defining the product profile for a new TB drug, before describing recent progress.