Accuracy in measuring intake of dietary constituents is an important issue in studies reporting the associations between diet and chronic diseases. We modified a Commonwealth Scientific and Industrial Research Organisation (CSIRO) food frequency questionnaire (FFQ) to include foods of interest in the field of Alzheimer’s disease (AD) research. The aim of the current study was to determine the reliability and validity of the AD-CSIROFFQ in 148 cognitively normal older adults. The AD-CSIROFFQ was completed before and after completion of a four-day weighed food record. Of the 508 food and beverage items reported, 309 had sufficient consumption levels for analysis of reliability. Of the 309 items, over 78% were significantly correlated between the two questionnaire administrations (Spearman’s rank correlations). We used two additional methods to assess absolute nutrient intake agreement between the AD-CSIROFFQ and the weighed food records (Pearson’s correlation coefficients and Bland–Altman plots) and quintile rankings to measure group level agreement. The adequate correlations observed between questionnaire responses suggest that the AD-CSIROFFQ is reliable. All nutrient intakes were acceptable for ranking of individuals on a group level, whilst the agreement levels with respect to the weighed food records for 11 of the 46 nutrients show validity in terms of their individual level absolute intake. The AD-CSIROFFQ makes an important contribution to the tools available for assessing usual dietary intake in groups of older adults with respect to AD research.