We propose assessment of MRS as an imaging biomarker of early Alzheimer's disease (AD) and its progression. It is widely recognized that biomarkers are important in the search for effective drug therapies for neurodegenerative disease. While amyloid imaging and CSF markers are useful in AD detection, their utility for progression of disease is less clear. MRI structural imaging is less invasive, but measures of atrophy may have limitations in studies of disease progression. It would be valuable to have an MRI-based method of assessing disease progression that might allow for monitoring of therapy, without depending on registration, mapping, or complicated measures of small brain volume losses. MRS is a biochemical measure that might be more sensitive than volume, and might also allow assessment of changes including those reversible. MRS can be used to assess neuronal (NAA signal normalized to Cr) and glial integrity (MI and CHO signals normalized to Cr). This exploratory proposal aims to perform cross-sectional study of 30 subjects, including 10 each of controls, MCI, and mild AD, determining regional changes in MRS signals including NAA, Chol, MI, and Cr, using a 2D planar acquisition technique. Analysis will compare at least four brain regions, including posterior cingulate cortex (PCC), premotor cortex (PMC), occipital cortex (OC), and parietal subcortical white-matter (PWM). Four methods will be used to increase sensitivity in this exploratory study: (a) Longitudinal change analysis, based on multiple scans performed at 0, 6, and 12 months; (b) Regional analysis, based on differential involvement of AD, which typically involves PCC more than PMC more than OC, more than PWM; (c) Severity analysis, in which MRS changes will be compared in cases of lesser and greater disease severity; and (d) biochemical analysis, based on measurement of NAA, Chol, MI, and Cr, to determine utility of different measures.