Abstract
The assessment of physical activity and energy needs is discussed with particular reference to elderly subjects. The only practical method of determining physical activity in epidemiological research remains a retrospective questionnaire. Attitude measurements may nevertheless help in assessing those individuals prone to exaggerate participation, while personal activity monitors and measurement of individual fitness levels may contribute to the refinement of questionnaires. Particular problems in applying questionnaires to elderly subjects can include poor vision, hearing, and recent memory, coupled with slow understanding and garrulousness. The design of forms should be as simple and unambiguous as possible, and timely. Attempts to distinguish occupational from leisure or household chores and to explore seasonal variations merely seem likely to cause confusion.