In 1993 the WHO declared tuberculosis a global emergency, and subsequently introduced the DOTS strategy, a technical and management package based on earlier work of the IUATLD and international experience with directly observed therapy. Despite successful implementation of most of the elements of this strategy in several African countries and settings, tuberculosis case rates continue to escalate where the prevalence of HIV infection is high. We explore possible reasons for the failure to control tuberculosis even in the context of tuberculosis programmes that have been considered models for others to emulate. In many African countries half or more of tuberculosis patients are now HIV-infected; in such settings, the overall epidemiology of tuberculosis is disproportionately affected by what happens in the HIV-infected subpopulation of the community. Persons with HIV infection are at increased risk of rapid progression following primary infection or re-infection, and also from reactivation of latent infection with Mycobacterium tuberculosis. More intensive strategies need to be targeted to the HIV-infected to interrupt on-going transmission (active and passive case detection; prevention of nosocomial transmission) and reactivation (preventive therapy). The high burden of other HIV-related disease in patients with tuberculosis, such as other bacterial infections, toxoplasmosis and other manifestations of AIDS, require that tuberculosis programmes integrate their activities better with those of HIV/AIDS programmes, including those for provision of HIV/AIDS care. Enhanced epidemiological surveillance is required to follow tuberculosis trends in the HIV-positive and negative sub-populations of communities, which may respond differently to control efforts. Strategies for tuberculosis control programmes in countries of high and low HIV prevalence cannot be the same, but must take into account the epidemiology of HIV infection. HIV/AIDS in Africa poses severe challenges of purpose and identity to tuberculosis control programmes, which have not adapted to the altered realities of the HIV/AIDS era. DOTS alone is unlikely to control tuberculosis in sub-Saharan Africa; one major achievement of DOTS when implemented, however, has been its apparent ability to limit the development and spread of drug resistance.