We investigated whether smoking is a risk factor for relative cognitive decline from age 11 to 64 years. The potentially confounding effects of childhood IQ, occupational status, level of education, presence of heart disease, hypertension, and lung function were examined. Subjects were nondemented and living independently. They were all born in 1936, had been participants in the same Scottish national IQ survey in 1947, and were reexamined at age about 64 years in 2000-2002. Current smokers and nonsmokers had significantly different mental test scores at age 64. This difference remained after adjustment for childhood IQ. Multiple linear regression identified childhood IQ, level of education, occupational code, lung function, and smoking history as significant independent predictors of mental function at age 64. In this sample, smoking makes a small (<1% variance) independent negative contribution to cognitive aging.