Tuberculosis (TB) is a disease of antiquity, caused by Mycobacterium tuberculosis, which principally affects the lungs. It is a major public-health problem, with around 9 million new cases and 2 million deaths estimated to occur each year. Patients with pulmonary TB whose sputum is smear-positive for M. tuberculosis form the main source of infection in communities. About 5%-10% of infected individuals are likely to develop symptomatic TB during their lives but the risk of developing the clinical manifestations of the disease is greatly increased by HIV co-infection. The strong association between HIV and TB in sub-Saharan Africa is responsible for the massive increase in the incidence of TB observed in that region in the last 20 years. Diagnosis of TB in resource-poor countries is largely based on sputum-smear microscopy and chest radiography, although these methods lack sensitivity or specificity, especially when used on HIV-infected patients. Effective treatment has existed for 40 years but TB-attributable mortality remains high among HIV-infected patients in Africa, who are also particularly likely to develop TB again after receiving drug treatment for the disease. In Eastern Europe it is drug resistance in the local M. tuberculosis that makes the treatment of TB relatively ineffective. The approach to TB control that is now internationally recommended is the DOTS ('directly-observed treatment, short-course') strategy, which aims to prevent the transmission of M. tuberculosis, and the related illness and death, by using combinations of anti-TB drugs to treat patients with the active disease. Unfortunately, countries in sub-Saharan Africa are falling short of the World Health Organization's targets for case detection and treatment. This failure is, in turn, making the achievement of the Millennium Development Goals for TB--to ensure that the incidence of TB is falling by 2015 and to halve the prevalence of TB and the annual number of TB-attributable deaths between 1990 and 2015--less likely. To improve the performance and impact of TB-control programmes, in the face of HIV co-infection and other constraints on DOTS, the World Health Organization has launched the revised 'Stop TB Strategy'. The new strategy, to be implemented via the Global Plan to Stop TB (2006-2015), includes intensified TB-case finding, treatment of latent TB infection with isoniazid, prevention of HIV infection, cotrimoxazole preventive therapy, and antiretroviral therapy.