Examining cognitive decline with aging will help us better characterize pathological and non-pathological changes in the brain throughout the lifespan. With the population over 60 rising rapidly, discoveries in this domain will likely have dramatic impact on public health. This dissertation investigates structural and functional changes in the medial temporal lobe memory system with aging. This work is based on converging insights from computational models as well as behavioral, electrophysiological, and neuroanatomical findings in rodent models of aging. The approach is based on the premise that the hippocampal dentate gyrus is critically involved in pattern separation, or the ability to discriminate new information from previously learned information. Recent work in the rodent suggests that agerelated degradation in perforant path input to the dentate gyrus and CA3 region from layer II entorhinal cortical neurons reduces the hippocampus' ability to successfully separate, resulting in shifting the computational bias away from new learning and towards previously stored information. The experiments presented in this dissertation provide compelling evidence in support of this model. Namely, we show that (1) aging results in behavioral impairments in pattern separation abilities in humans, (2) aberrations in functional activity in the CA3/dentate gyrus region measured using BOLD fMRI mechanistically contribute to this deficit, and (3) age-related degradation in the perforant path is the underlying structural basis for functional and behavioral degradation in pattern separation in humans. This work represents an important step towards bridging the gap between animal work and human work and translates into a powerful tool for understanding what differentiates healthy from pathological aging.