Two follow-up studies were designed to analyze die cross-seciional and longitudinal associations between cognitive functioning and physical activity among two cohorts of elderly people. At baseline, over 90% of the 75- and 80-year-old populations were interviewed at home and almost 80% participated in the laboratory examinations. Cognitive functioning was assessed by psychometric tests and reaction time tasks, and physical activity was assessed by a subjective self-assessment as well as by objectively measured maximal walking speed. Among both cohorts, the decline over the 5-year period in cognitive functioning as well as in physical activity was generally small but statistically significant. The test-retest correlations were higher for the cognitive functioning scores than for the physical activity variables. The associations between cognitive functioning and physical activity were inconsistent and showed some differences between men and women.