Researchers have suggested that educational attainment and occupational status—indicators of cognitive and/or neurologic “reserve”—can help persons compensate for clinical manifestations of Alzheimer's disease (AD), such as the rates of cognitive and functional decline. The effects of educational attainment on rates of decline could be “direct” (independent of occupational status), “indirect” (working through occupational status), or both. We used multilevel analysis for repeated measures to study the effects of educational attainment and occupational status on rates of decline in cognition (Mini-Mental State Examination, MMSE) and function (Cleveland Scale for Activities of Daily Living). Subject included persons with “probable” or “possible” AD, drawn from our Alzheimer's Disease Research Center registry (N = 482 in the analysis of cognitive decline, and N = 450 in the analysis of functional decline). When controlling for year of birth, gender, ethinicity, and duration of illness, we found that there was an inverse relationship between number of years of education and rate of decline in MMSE, but effects of occupational status were not significant. This implies a “direct” effect of education on decline in MMSE, but no “indirect” effect through occupational status. Neither educational attainment nor occupational status affected rate of decline in functional ability. We conclude that education slows the rate of cognitive decline in persons with AD, but not through its impact on occupational status. Thus the protective effects of reserve may be established early in life, before people enter the workforce.