The neuropathogenesis of HIV-1 infection of the central nervous system (CNS) has been intensively studied, yet it remains incompletely understood. The virus enters the CNS early after systemic infection, establishes a viral reservoir, and remains largely inaccessible to antiretroviral therapies. In certain individuals, slowly evolving cognitive dysfunction becomes manifested years after infection, resulting in frank dementia (AIDS dementia complex [ADC]) in approximately 10% of individuals, with less severe cognitive dysfunction in up to 50% of all infected individuals. Neuropathologic and neuroimaging studies demonstrate that reversible metabolic disturbances occur in neurons and glia early after CNS invasion and that progressive, irreversible neurodegeneration may precede overt clinical deterioration by years. In vitro studies demonstrate that HIV-1 infection of macrophages and microglia, the endogenous brain macrophages, releases soluble factors that induce neuronal cell death. The viral and host factors that modulate neuronal cell death in vivo are only partly understood. Evolving therapies directed toward preventing neurodegeneration in HIV infection include CNS-penetrating antiretroviral drugs and pharmacologic agents that protect neurons from effects of HIV-1-infected macrophage neurotoxins. Major unanswered questions concerning the pathogenesis and prevention of HIV-1-induced CNS degeneration include (1) which cellular and viral factors are primarily responsible for HIV-1-induced neurodegeneration in vivo; (2) which surrogate markers of HIV-1 infection in the CNS (eg, viral load or cytokines) can predict risk for neurodegeneration; and (3) can antiretroviral drugs or specific CNS-targeted agents protect against HIV-1?.