This chapter describes the different methods used to perform magnetic resonance spectroscopy (MRS) studies of patients with HIV infection, reviews the major findings from the studies, and identifies future directions for clinical studies that may further elucidate the pathophysiology of HIV-associated brain injury, as well as discusses issues to consider in treatment monitoring. MRS studies have been performed in both adult and pediatric patients with HIV infection, using localized MRS and MRS imaging (MRSI) techniques and long TEs and short TEs, and before or after antiretroviral treatment. The majority of the studies used localized spectroscopy techniques (either point resolved spectroscopy (PRESS) technique or the stimulated echo acquisition method) and evaluated one to three brain regions. Limited by magnetic-susceptibility problems in evaluating the frontal brain regions, many of the earlier localized MRS studies evaluated the parietal or the occipital brain regions. Improvements in MRS techniques, such as adjustments of the slice order, allow the assessment of the frontal lobe and subcortical brain regions. Prior to highly active antiretroviral therapy (HAART), several longitudinal MRS studies reported changes in metabolite ratios or levels in HIV patients during antiretroviral treatment. Levels of inflammatory markers, such as the chemokine macrophage chemoattractant protein 1 (MCP-1), were found to be higher in patients with HIV-associated dementia than in those who were neuroasymptomatic. Additionally, both in vivo and ex vivo MRS studies in animal models of AIDS (e.g., simian immunodeficiency virus (SIV) and feline immunodeficiency virus (FIV)) should provide additional insights into the pathophysiology of HIV-associated dementia.