Recent advances in studies of HIV and SIV pathogenesis have renewed interest in the role of cellular activation in mediating disease. A critical finding was that “natural hosts,” i.e., non-human primates that have little if any disease associated with chronic lentivirus infections, have high and sustained viremia. Parallel studies have also demonstrated that the gut immune system (GALT) is significantly depleted in these natural hosts. Yet, in contrast with humans infected with HIV, in nonhuman primates, natural infections do not appear to be a serious compromise of the mucosal barrier to microbial translocation. In HIV-infected people, depletion of GALT (with perhaps a preference for TH17 cells) results in abnormally high levels of lipopolysaccharides (LPS) and is associated with higher numbers of T cells expressing activation markers. This may influence T cell turnover and depletion. At the same time, circulating monocytes express higher levels of the activation marker CD16⁺ in association with higher LPS levels. In theory, this could promote entry into organs, such as the central nervous system, where activated cells have long been recognized as playing an important role in mediating neuropathology. However, while elevated circulating LPS levels are indeed concordant with a higher frequency of HIV-associated dementia, the relationship between CD16⁺ cells and dementia is less clear. Whether ARV completely abrogate neurocognitive problems is also still open to interpretation. Other recent experiments have emphasized the key role of the mucosal tissue in the initial infections with lentiviruses, including the potential role of a local inflammatory and apoptotic reaction in promoting initial local replication and eventual spread, which nevertheless provides a significant bottleneck so that a very limited subset of viruses actually initiate the systemic infection.