Although highly active antiretroviral therapy (HAART) has been successful in suppressing plasma HIV RNA levels in infected patients, it has not resulted in eradication of virus. It is now clear that virus replication persists despite undetectable plasma viremia in individuals receiving HAART. In this regard, withdrawing HAART, even after prolonged periods of virus suppression, leads to an almost universal rapid rebound of plasma viremia. It is also now clear that prolonged, continuous HAART carries a risk of significant toxicity and side effects. These recent observations may argue for a different approach to HAART with the goals of: 1) durable suppression of virus replication, without an attempt at eradication, and; 2) minimization of toxicity and side effects and improvement in patient life-style. Therefore, we propose to study the virologic and immunologic effects of intermittent versus continuous HAART in HIV-infected individuals as a possible means to achieve these goals. Our primary interest is to determine if 30-70% less HAART than is required for continuous therapy, administered in a cyclic, intermittent fashion, can effectively maintain suppression of HIV replication and CD4+ T cell counts to levels equivalent to continuous treatment. These studies include a randomized, controlled trial of long cycle intermittent HAART (1 month off drugs followed by 2 months on HAART for 22months) and several pilot studies of short cycle intermittent HAART (e.g. 7 days off HAART followed by 7 days on HAART for up to 2 4 months). The reduction in medication needs will reduce cost and may slow or prevent the onset of toxicity while enhancing adherence.