Dietary fibre has a complex and highly variable composition. Although some dietary fibres may protect against colorectal cancer, it is unlikely that all are equally protective. Dietary fibre is principally composed of plant cell walls, but it also includes components obtained from cell walls (e.g. cellulose, pectin, and lignin), and non-starch polysaccharides (NSPs) from other sources (e.g. seaweeds and micro-organisms). The AOAC and Englyst methods are commonly used to determine the total amount of dietary fibre in foods. Most of the cell walls in food plants are from parenchyma cells, which are extensively degraded by bacteria in the colon. Cell types with walls containing the hydrophibic polymers lignin, suberin, or cutin also occur in food plants in small numbers, but they may be important in preventing colorectal cancer. Lignin, and possibly the other polymers, protect these walls from degradation. Epidemiological, human intervention, and animal studies can be used to try to identify the most protective dietary fibres. Epidemiological studies are difficult to interpret because usually only the total amount of dietary fibre eaten is reported. Intervention studies indicate that wheat bran dietary fibre may be protective. The results of animal carcinogenesis studies are variable, but sources of insoluble dietary fibres, including wheat bran, appear more protective than soluble dietary fibres, and some dietary fibres appear to enhance carcinogenesis. Possible mechanisms for protection by dietary fibres can be divided into two groups: those where the dietary fibre is acting directly, and those which result from the dietary fibre being degraded by colonic bacterial enzymes and the products fermented. Possible direct mechanisms include the binding of carcinogens to undegradable dietary fibres, and the absortion of water by undegradable dietary fibre resulting in increased faecal bulk and shortened transit times. Possible indirect mechanisms include the lowering of the colon pH by the short-chain fatty acids produced by bacterial fermentation, and the specific effects of butyrate. There are also a number of possible mechanisms by which some dietary fibres may enhance carcinogenesis. Use of better defined dietary fibres will increase our understanding of the role of dietary fibres in modulating colorectal cancer.