Before highly active antiretroviral therapies (HAART) were available for the treatment of persons with HIV infection, disseminated Mycobacterium avium-intracellulare complex (MAC) infection was one of the most common opportunistic infections that affected people living with AIDS. Routine use of chemoprophylaxis with a macrolide has been advocated in guidelines for the treatment of HIV-infected individuals if they have a circulating CD4+ cell count of ≤50 cells/μL. In addition, lifelong prophylaxis for disease recurrence has been recommended for those with a history of disseminated MAC infection. The introduction of HAART has resulted in a remarkable decline in the incidence of opportunistic infections and death among persons living with AIDS. Considerable reconstitution of functional immune responses against opportunistic infections can be achieved with HAART. In the case of infection with MAC, there has been a substantial reduction in the incidence of disseminated infections in the HAART era, even in countries where the use of MAC prophylaxis was never widely accepted. Moreover, the clinical picture of MAC infections in patients treated with potent antiretroviral therapies has shifted from a disseminated disease with bacteraemia to a localised infection, presenting most often with lymphadenopathy and osteomyelitis. Data from several recently conducted randomised, double-blind, placebo-controlled trials led to the current practice of discontinuing primary and secondary prophylaxis against disseminated MAC infections at stable CD4+ cell counts >100 cells/μL. These recommendations are still conservative as primary or secondary disseminated MAC infections are only rarely seen in patients who respond to HAART, despite treatment initiation at very low CD4+ cell counts. Potential adverse effects of macrolide therapy and drug interactions with antiretrovirals also metabolised via the cytochrome P450 enzyme system must be critically weighed against the marginal benefit that MAC prophylaxis may provide in addition to treatment with HAART. These authors feel that, unless patients who initiate HAART at low CD4+ cell counts do not respond to HIV-treatment, routine MAC prophylaxis should not be recommended. Nevertheless, the patient population for whom MAC prophylaxis may still be indicated in the era of HAART needs to be identified in prospectively designed clinical trials.