Background: HIV-1 Gag protein selectively packages two copies of the viral genome, an event thought to be primarily mediated by interactions between the packaging signal (Ψ) on the viral genome and the nucleocapsid domain of Gag. However, the Ψ sequence is defined based only on genetic studies and limited in vitro binding data. To date, no assay has been able to demonstrate a direct and specific interaction between Ψ and Gag protein in a relevant context, i.e. in live cells and in virions. Importantly, deletion of Ψ does not completely abolish genome encapsidation, suggesting that packaging may be more complex than Ψ:Gag interaction, and other regions on the viral genome may contribute. Finally, Gag undergoes several changes in localization, multimerization state and is proteolytically processed during particle genesis, but it is completely unknown how these changes affect RNA-binding properties of Gag. Therefore, we set out to determine globally and at near-nucleotide resolution the RNA targets of Gag during various stages of virion genesis, in cells and in virus particles. Methodology: To identify the RNA targets of Gag, we adapted a CLIP-seq (crosslinking-immunoprecipitation-sequencing) methodology, which combines immunoprecipitation of covalently crosslinked protein-RNA complexes with high-throughput sequencing. This method globally identifies the RNA molecules associated with an RNA-binding protein of interest in biological settings and provides near-nucleotide-resolution information about the protein-RNA interaction sites. Results: In addition to determining precisely where within Ψ Gag is bound, we identified several novel sites on the viral genome that are specifically bound by Gag in the cytosol of cells. Mutation of these sites delayed virus replication and we are currently investigating whether this is due to defects in genome packaging. Experiments performed on fractionated cells and in virions indicated that there are major changes in RNA binding specificity during particle morphogenesis: in contrast to cytosolic Gag, Gag associated with the plasma membrane and in virions binds to numerous other regions on, but does not uniformly coat, the viral genomic RNA. We have also globally identified several cellular RNA molecules, such as 7SL RNA, that are specifically bound by Gag in the cytosol and encapsidated into virus particles. Conclusions: Our studies indicate that Gag selects viral genome for packaging by binding to a few distinct regions on the viral genome and that association with the plasma membrane and particle formation induces significant changes in the RNA-binding properties of Gag. These studies provide the first dynamic, quantitative and high-resolution picture of viral/cellular RNA interactions with Gag in relevant biological contexts.