A typical aspect of aging is the increase in variability of a feature within a population, suggesting individual variation in the decline with age. The decline is probably inevitable (predisposed), but some show slower declines than others. Physical activity might be one of the avenues to influence the aging process. There is still a need for validated techniques of estimating habitual physical activity to study the relationship between physical activity and health. Doubly labeled water studies show a decrease in the average proportion of daily energy expenditure for physical activity of about 35% at the age of 20 to 25% at the age of 90. In absolute figures, the activity-induced energy expenditure falls to one third of the initial value over the corresponding age interval. Exercise training does not seem to prevent the age-associated decline in physical activity because of compensation by a decrease in non-training physical activity, in contrast to younger age groups. The effects of the habitual activity level and of exercise on the age-related decline in muscle mass are small if present at all. There is, however, a clear-cut effect of habitual activity and exercise training on muscle function. The positive effects are reflected in muscle fiber type, capillary density, and aerobic capacity. Physical inactivity is an important determinant of disability and mortality risk. Even a recent increase in activity level can have an important effect, although the benefits are also easily lost when the activity level is again reduced.