The purpose of this review is to articulate how progress in epidemiological research on physical activity and cancer prevention can be made. This report briefly reviews the accumulated evidence for an etiological role of physical activity in the prevention of cancer of the colon, breast, prostate, testes, lung, endometrium, and ovary and summarizes the evidence for a causal association for each of these sites. The evidence for a causal association between physical activity and colon and breast cancers is found to be "convincing," for prostate cancer to be "probable," for lung and endometrial cancers to be "possible," and for testicular and ovarian cancers to be currently "insufficient" to make any definitive conclusions. The emerging literature on physical activity and cancer prevention intervention studies is presented, and an overview of the literature on physical activity intervention is also provided. Given the level of evidence that is currently available for the associations between physical activity and cancer, it is argued that for additional progress to be made in this field, there need to be intervention studies on physical activity and cancers of the colon and breast. For the remaining cancer sites, better designed observational epidemiological studies are needed that address the identified methodological limitations found in previous studies. These limitations include crude and incomplete physical activity assessment, lack of adequate control for confounding and effect modification, as well as a lack of consideration of the underlying biological mechanisms that are operative. This review concludes with detailed recommendations for future research in this field.