Several studies suggest that glucose hypometabolism may be present in specific brain regions in cognitively normal older adults and could contribute to the risk of subsequent cognitive decline. However, certain methodological shortcomings including a lack of partial volume effect (PVE) correction or insufficient cognitive testing confound the interpretation of most studies on this topic. We combined ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET) and magnetic resonance (MR) imaging to quantify cerebral metabolic rate of glucose (CMRg) as well as cortical volume and thickness in 43 anatomically-defined brain regions from a group of cognitively normal younger (25±3 y old; n=25) and older adults (71±9 y old; n=31). After correcting for PVE, we observed 11-17% lower CMRg in three specific brain regions of the older group - the superior frontal cortex, caudal middle frontal cortex, and the caudate (p≤0.01 false discovery rate-corrected). In the older group, cortical volumes and cortical thickness were 13-33% and 7-18% lower respectively in multiple brain regions (p≤0.01 FDR correction). There were no differences in CMRg between individuals who were or were not prescribed antihypertensive medication. There were no significant correlations between CMRg and cognitive performance or metabolic parameters measured in fasting plasma. We conclude that highly localized glucose hypometabolism and widespread cortical thinning and atrophy can be present in older adults who are cognitively normal as assessed using age-normed neuropsychological testing measures.