Evidence for an independent role of increased physical activity in the primary prevention of coronary heart disease has grown in recent years. The authors apply the techniques of meta-analysis to data extracted from the published literature by Powell et al. (Ann Rev Public Health 1987;8:253-87), as well as more recent studies addressing this relation, in order to make formal quantitative statements and to explore features of study design that influence the observed relation between physical activity and coronary heart disease risk. They find, for example, a summary relative risk of death from coronary heart disease of 1.9 (95% confidence interval 1.6-2.2) for sedentary compared with active occupations. The authors also find that methodologically stronger studies tend to show a larger benefit of physical activity than less well-designed studies.