This prospective study investigated the association between smoking and cognitive performance in a community of nondemented elderly subjects aged 65 or older. All subjects were categorized as current smokers, former smokers, or never smokers. The lifetime cigarette exposure was computed. At baseline, we found the abstainers from smoking had better cognitive performances; however, the differences were not significant after adjusting for age, education, hypertension, diabetes, and vascular events. The lifetime cigarette exposure was not predictive of the cognitive status. At a 3-year follow-up, neither the smoking status nor the lifetime cigarette exposure predicted the declination of cognition.