Evidence that physical activity may protect against various forms of cancer is examined in relation to occupational demands, leisure activities and participation in sport while at university. The variety of forms of neoplasm and equally varied physical activity histories militate against finding any simple relationship between the risk of malignancy and the individual's physical activity history. Nevertheless, five of seven major occupational studies suggest that a physically active occupation offers some protection against colon cancer, and an application of Bradford Hill's criteria generally supports the causal nature of the relationship between physical inactivity and an increased risk of intestinal neoplasia. However, existing reports are by no means conclusive; there thus remains a need for well-designed epidemiological studies of this issue. Data from one laboratory also suggest that in women a history of active leisure is associated with a reduced prevalence of breast and reproductive system cancers. Physical activity potentially encourages a healthy lifestyle, and it could have more direct effects on certain forms of carcinogenesis (for instance, by a speeding of gastro-intestinal transit, or a moderation of sex hormone levels). However, there are also potential negative effects from some types of exercise, particularly an excessive exposure to ultra-violet light in certain water sports. Since moderate exercise elevates mood and helps to conserve lean tissue, it may finally be a helpful component of treatment after a neoplasm has been diagnosed.