Although highly active antiretroviral therapy (HAART) decreases HIV-associated mortality, it does not to completely restore health. Patients doing well on otherwise effective HAART remain at risk for cancer, cardiovascular/liver disease, osteopenia, and other "non-AIDS-defining" events. While complete eradication may never be feasible, a "functional cure" in which patients are able to maintain undetectable viral loads indefinitely without therapy may be possible. The best evidence for this are the so-called "elite" controllers, whom we define as individuals who are HIV-seropositive, with plasma HIV RNA levels below the level of conventional detection without treatment. Controllers may be conceptualized as a naturally occurring model of a functional cure (or "HIV remission"), and are ideal patients in which to study HIV persistence and the possibility of eradication. We propose to conduct a pilot study to better characterize the reservoirs that lead to viral persistence in a group of well-characterized controllers. We propose two specific aims: 1) to characterize the dynamics of viral production in blood and gut-associated lymphoid tissue (GALT) in controllers; and 2) to prospectively treat 10 controllers with raltegravir, tenofovir/emtricitabine for 24 weeks and study the effects of HAART on viral dynamics and host inflammatory responses. Our primary hypotheses are: 1) viral replication is ongoing in untreated controllers, 2) HAART will reduce viral replication in blood and GALT and decrease immune activation, and 3) higher levels of immune activation are associated with greater measures of microbial translocation and distribution of virus to more differentiated T cell subsets.