The Scottish Mental Surveys of 1932 and 1947 collected valid IQ-type test scores for almost everyone born in 1921 and 1936 and attending school on June 1, 1932 (N=89,498) and June 4, 1947 (N=70,805). These surveys are described. This research, using the surveys' data, examined (a) the stability of intelligence differences across the life span, (b) the determinants of cognitive change from childhood to old age, and (c) the impact of childhood intelligence on survival and health in old age. Surviving participants of the Scottish Mental Surveys were tested, and the surveys' data were linked with public and health records. Novel findings on the stability of IQ scores from age 11 to age 80; sex differences in cognitive aging; the dedifferentiation hypothesis of cognitive aging; and the effect of childhood IQ on all-cause and specific mortality, morbidity, and frailty in old age are presented.