Education, occupation, premorbid intelligence, and brain size are surrogate markers for cognitive reserve. Whether these markers have biological influence on Alzheimer disease (AD) pathology is not known. We thus aimed to investigate the effect of cognitive reserve proxies on longitudinal change of AD biomarkers. A total of 819 participants with normal cognition, mild cognitive impairment, and mild AD were enrolled in the Alzheimer's Disease Neuroimaging Initiative and followed up with repeated measures of cerebrospinal fluid, positron emission tomography, and magnetic resonance imaging biomarkers. Generalized estimating equations were used to assess whether biomarker rates of change were modified by reserve proxies. Cerebrospinal fluid Aβ42 decline was slower in normal cognition participants with higher cognitive reserve indexed by education, occupation, and American National Adult Reading Test (ANART). The decline of [F] fluorodeoxyglucose positron emission tomography uptake was slower in AD participants with better performance on the ANART. Education, occupation, and ANART did not modify the rates of magnetic resonance imaging hippocampal atrophy in any group. These findings remained unchanged after accounting for APOE 4, longitudinal missing data, and baseline cognitive performance. Higher levels of reserve markers may slow the rate of amyloid deposition before cognitive impairment and preserve glucose metabolism at the dementia stage over the course of AD pathologic progression.