Understanding of human immunodeficiency virus (HIV) neuropathogenesis has been greatly enhanced by careful analyses of postmortem brain tissue, as well as by primate models of infection with related lentiviruses. However, brain tissue from persons living with HIV infection is rarely available for study, and elegant observations from primate systems may require confirmation in humans. Because of these inherent limitations, it is important to learn as much as possible by studying cerebrospinal fluid (CSF) during HIV infection. The present discussion considers selected issues relevant to the study of CSF from HIV-infected individuals. These include a strategy to intensely sample CSF to better understand viral dynamics and the role of the brain as a pharmacologic sanctuary site, evidence for HIV sequence diversity in CSF and plasma during HIV therapy and the implications of such diversity, and the importance of host genetics relevant to studying HIV neuropathogenesis.