During the last 2 decades, substantial progress has been made in understanding the relationship between dietary constituents and the development of colon cancer in man. Unlike studies of cancer among smokers and nonsmokers, nutritional epidemiologic studies are confronted with the inherent difficulty of assessing reasonably precise exposures. The lack of consistency between international correlation studies and case-control studies does not necessarily negate a dietary etiology of colon cancer because these inconsistencies may have arisen, at least in part, from methodological limitations. Some of these deficiencies in epidemiological studies of diet and cancer have been corrected; recent case-control studies demonstrated that high dietary fat is a risk factor for colon cancer development and that an overall increase in intake of foods high in fiber might decrease the risk for colon cancer. The results of epidemiologic studies may be assumed to present conservative estimates of the true risk for cancer associated with diet. The populations with high incidences of colon cancer are characterized by high consumption of dietary fat, which may be a risk factor in the absence of factors that are protective, such as whole-grain cereals and of other high fiber. Laboratory-animal model studies have shown that certain dietary lipids and fibers influence tumorigenesis in the colon. The data of metabolic epidemiological and laboratory-animal model studies are sufficiently convincing with respect to the enhancement of colon cancer by type of fat and protection by certain dietary fibers.