A category effect (i.e., living vs. nonliving exemplars) in confrontation naming has been reported in association with various cerebral pathologies. However, the published reports conflict as to the presence of this category effect in normal controls. The present experiment included 90 subjects in three age groups (young, young-elderly, and old-elderly) and controlled the stimuli in two major ways: by increasing the difficulty level to avoid a ceiling effect, and by equating the stimulus lists on five important word property variables (word frequency, familiarity, age of acquisition, visual complexity, and name agreement). The results confirmed that the category effect was present in the two elderly groups but not in the younger group. However, a gender effect was evidenced in the younger group but not in the two elderly groups. Overall, these results suggest that the category effect reported in neurological populations represents at least in part the influence of normal aging on semantic memory.