The application of epidemiological methods to exercise immunology is reviewed briefly, with particular reference to the possible influences of physical activity, exercise and training on susceptibility to upper respiratory infections. Available reports are arbitrarily rated in terms of limiting factors: the quality of the assessment of physical activity, the precision of diagnosis of upper respiratory infection and overall methodology. The pattern of physical activity has often been clearly established but, in part because of the problems associated with the competitive environment, assessments of infection and overall methodology have often been less than optimal. Although there is some evidence that susceptibility to infection is increased by either a single bout of very heavy activity or a period of heavy training, reports are far from unanimous, and in certain respects fail to meet the classical epidemiological criteria of a causal relationship. The issue is important to both the health and the success of the international competitor, and merits definitive investigation, using optimal methods to assess both activity patterns and infection.