Alzheimer’s disease (AD) is the most common form of dementia with hallmarks of ß-amyloid (Aß) plaques, tau tangles, and neurodegeneration. Studies have shown that neurodegeneration components, especially brain metabolic deficits, are more predictable for AD severity than Aß and tau. However, detailed knowledge of the biochemical composition of AD brain tissue vs. normal brain tissue remains unclear. In this study, we performed a metabolomics analysis on the brain tissue of 158 community-based older adults in the University of Kentucky AD Research Center brain bank to characterize the biochemical profiles of brains with and without AD based on white/gray matter type, apolipoprotein E genotype (e3 vs e4 variants), and disease stage (early vs late) as all these factors influence metabolic processes. We also used machine learning to rank the top metabolites separating controls and AD in gray and white matter. Compared with control samples, we found that glutamate and creatine metabolism were more critical for predicting AD in the gray matter, while glycine, fatty acid, pyrimidine, tricarboxylic acid (TCA) cycle, and phosphatidylcholine metabolism were more critical in the white matter. In e4 carriers, metabolites associated with the TCA cycle and oxidative phosphorylation were prominent in advanced stages compared to the early stages. In e3 carriers, metabolites related to oxidative DNA damage, changes in inhibitory neurotransmitters, and disruptions of neuronal membranes were prominent in advanced stages compared to the early stages. In early disease, e4 carriers had metabolites related to poor kidney function and altered neuronal sterol metabolism compared to e3 carriers, but there were few differences between genotypes in late disease. Our results indicate that metabolism plays a pivotal role in differentiating APOE- and stage-dependent changes in AD and may facilitate precision lifestyle and dietary interventions to mitigate AD risk in the early stages, especially for e4 carriers.