Background: Gray matter atrophy is regarded as a valid marker of neurodegeneration in Alzheimer's disease (AD), but few studies have investigated in detail the topographic changes associated with normal aging. In addition, few studies have compared the changes in the earliest clinical stage of AD (prodromal AD (pAD)) with those of healthy aging. Here we aimed to investigate the topographical distribution of age-related cortical atrophy and to compare it with that associated with prodromal and estabilished AD. Methods: Structural T1-weighted high-resolution brain magnetic resonance imaging scans were acquired from 60 healthy volunteers (20 young adults, YA: age 32.7 ± 4.5 years; 40 elderly subjects, HE: age 71.3 ± 6.2 years), 16 mild cognitive impairment subjects who converted to AD within 2 years (prodromal AD, pAD: age 72.8 ± 5.4), and 20 mild to moderate AD patients (mAD, age 72.5 ± 10.3). Cortical gray matter differences were investigated using a surface-based anatomical mesh modeling technique (cortical pattern matching) and region-of-interest (ROI) analyses based on hypothesized brain networks taught to have a functional and a structural link to each other. Differences in cortical atrophy were assessed between groups, as well as the effect of age within groups. Results: HE compared to YA showed a 10-30% deficit in cortical gray matter in widespread frontal, temporal, and parietal regions (p = 0.0001 by permutation testing), 6-13% loss in the visual and sensorimotor cortices (p <0.01) and up to 13% loss in the direct hippocampal pathway ROIs (p <0.001). pAD patients showed on average 8-9% cortical loss compared to HE (p <0.0001), mainly in the left (up to 6% loss, p = 0.06) and right polysynaptic hippocampal pathway ROIs (up to 8% loss, p = 0.01), and in the left and right olfactory/orbitofrontal cortex (up to 12-15% loss, p <0.001). The pattern of cortical atrophy in mAD versus HE was similar to that in pAD, but was more severe in the direct hippocampal pathway ROIs and sensorimotor, visual and temporal cortices (13-15% loss compared with HE, p <0.0001). Conclusion: Gray matter loss occurs during aging with rates of atrophy even more severe than that observed during the course of AD. These changes may be caused by normal mechanisms. In pAD, cortical atrophy due to disease is milder than that due to aging, maybe resulting from a slowed down velocity of cell loss, but affects specific brain areas. These findings are consistent with the view that AD is not merely accelerated aging.