Several epidemiological studies link consumption of fibre-rich foods to a reduced risk of type 2 diabetes and CHD. The ‘fibre hypothesis’ suggested that this was a direct effect of fibre. However, fibre-rich foods contain different types of fibre as well as other potentially beneficial compounds, and many foods naturally high in fibre have low glycaemic and insulinaemic indices, possibly due to food form. The question therefore emerges as to the effect of isolated fibre per se on insulin sensitivity, lipids and other risk factors associated with the metabolic syndrome. Many beneficial effects are seen with pharmacological doses of isolated viscous soluble fibre, including improved insulin sensitivity, decreased LDL-cholesterol levels and decreased clotting factors. Similar effects are seen with low glycaemic-index foods. In contrast, insoluble non-viscous cereal fibre is not seen to act directly on risk factors when taken in refined foods such as in milled flour. Since cereal fibre, the major type of fibre in western diets, does not directly act on the risk factors for the metabolic syndrome, the question remains as to possible mechanisms. Until now, fibre and the nature and processing of the starch and particle size have been seen as the main determinants of the metabolic response to starchy foods. However, fibre-rich foods also have an increased protein-to-carbohydrate ratio. Hence we suggest that the protective effect of fibre may also be due to increased vegetable protein content, which may act directly to reduce clotting factors and oxidized LDL-cholesterol levels.