HIV establishes a latent reservoir in a small pool of resting CD4+ T cells early in the infection of a new host. This viral reservoir has a very slow rate of decay and it is resistant to anti-retroviral therapy and the immune surveillance. This reservoir poses a significant obstacle to the eradication of virus in an infected individual. It is thus of importance to understand how this reservoir is established and what are the requirements for its establishment. One widely accepted theory suggests that latently infected resting cells arise as activated cells become infected during the transition to a resting state. Contrary to this model, the work done for this dissertation supports an alternative model for the establishment of the latent reservoir. In this alternative model, HIV can establish latency in resting CD4+ T cells by direct infection. Using sensitive kinetic PCR techniques and careful analysis of HIV DNA intermediates at late time points post- infection in vitro, we found that HIV can fuse, reverse transcribe and finally integrate directly into resting CD4+ T cells in absence of cellular activation. The findings presented here suggest that this process can take place even when the resting cells are exposed to low inoculums of virus. In addition, work in this dissertation indicates that HIV can integrate directly into both major subsets of resting CD4+ T cells, memory and naïve cells. Therefore, both memory and naïve resting CD4+ T cells can serve as latent reservoirs for HIV. Finally, the finding that HIV can integrate in resting cells suggests that gene therapy approaches that target resting CD4+ T cells can be developed. Altogether, the findings in this dissertation have the potential to influence the focus of future therapeutic drug approaches for the reduction or elimination of the latent reservoir and suggest that alternative anti-retroviral approaches, such as gene therapy of resting cells, are possible.