Highly educated individuals have a lower risk of developing dementia and Alzheimer’s disease (AD). A common assumption is that their “cognitive reserve” protects them from cognitive decline and postpones the clinical manifestation of dementia. These highly educated individuals usually obtain normal scores on cognitive screening tests, although at the same time they can experience subjective cognitive decline and difficulty in multiple cognitive domains. Although comprehensive neuropsychological evaluations usually identify subtle changes in cognition, they demand extensive resources and thus are expensive and difficult to obtain. Therefore, lack of sensitivity of screening tests on the one hand, along with difficulty to acquire a comprehensive neuropsychological evaluation on the other hand, impede identification of cognitive decline at its earliest stages in this special population. Accordingly, this study aims to identify which neuropsychological tests have the highest sensitivity to detect the earliest stages of cognitive decline among highly educated elderly [n = 27, ages 66–80 (mean = 72.6 SD = 4.54), mean education level = 17.14 (SD = 3.21 range: 12–24 years)]. Baseline scores and scores at one-year follow up were obtained. We also conducted MRI scans to characterize the relation between brain volume and cognitive performance. Results show significant reductions in RVALT, Semantic verbal Fluency, ROCF copy, and MoCA scores whereas PF, TMT, ROCF delay, digit span, and knowledge tests were not significant. The study stresses the importance of using sensitive neuropsychological tests to examine this special population and the need to create norms that combine an individual’s education with age.