Recent research on the prevalence of age-associated memory impairment (AAMI) has reflected considerable variability, with estimates ranging from 35% to 98%. This variability is attributed to (a) failure to employ the complete diagnostic criteria for AAMI and (b) failure to consider age as a variable in estimating prevalence. Analysis of published normative data on both standard clinical memory tests and computer-simulated everyday memory tests shows a clear increase in the percentage of persons meeting the AAMI memory performance criterion as a function of age. These data are offered as an upper-bound estimate of the prevalence of AAMI, by age decade.