A systematic review of prospective cohort and case-control studies investigating relationships between the intake of fish and incidence of prostate, breast, or colorectal cancers was conducted. A total of 106 studies fulfilled the requirements stated in the "Search strategy and selection criteria." Among 273 estimates of association reported by these studies, 53 indicated decreased risk while 12 indicated increased risk associated with fish intake. The hypothesis linking fish consumption and low cancer incidence appears to be supported by little epidemiological data. However, there are several factors that may mask potential protective associations with fish intake. The type and the amount of fish eaten, the cooking method, the stage of the cancer and, amongst women, menopausal status may all be important factors that relate to whether fish is protective or not. Future epidemiologic studies with a clearer assessment of these factors are needed to know more about the effects of fish consumption on cancer risk. Therefore, until there are better measures of dietary exposure or biomarkers to correlate self-report, no conclusion can be drawn regarding the recommendation for increasing fish consumption in general to reduce the risk of developing the most common cancers in Western societies.