Ergonomic epidemiology is a rapidly increasing field of research providing data on the occurrence of musculoskeletal disorders and possible risk factors. The present paper states, on the basis of a literature overview, that physical work load (mechanical exposure) is poorly defined and measured in most studies on ergonomic epidemiology. On this background the paper: (1) suggests adjustments of mechanical exposure concepts and terminology; (2) concludes that invalid exposure assessment may, to a large extent, explain the lack of quantitative data on relationships between mechanical exposures and musculoskeletal disorders; and (3) suggests some guidelines for future quantitative assessments of mechanical exposure in large populations.