Epidemiology and animal experiments indicate that dietary fibres protect against the development of colorectal cancer. However, insoluble dietary fibres appear to be more effective than soluble dietary fibres and one mechanism by which they may protect is by adsorbing dietary carcinogens. We found previously that the ability of a carcinogen to adsorb in vitro to alpha-cellulose (a model insoluble dietary fibre) was strongly related to the hydrophobicity of the carcinogen, measured as the calculated logarithm of the partition coefficient between 1-octanol and water (C log P). Furthermore, soluble dietary fibres (soluble-fibre polysaccharides), including gum arabic, reduced the adsorption of the hydrophobic carcinogen, DNP, to alpha-cellulose. In the present study we tested the ability of gum arabic to reduce the adsorption in vitro of the carcinogens BaP (C log P = 6.124), DNP (C log P = 4.384), and the heterocyclic amines, Trp-P-1 (C log P = 3.230) and MeIQx (C log P = 1.078). Gum arabic reduced the adsorption to alpha-cellulose of BaP and DNP, but not the adsorption of Trp-P-1 or MeIQx. Gum arabic also reduced the adsorption of BaP to an insoluble, dietary-fibre preparation from commercial cork which contains the hydrophobic component, suberin, but did not affect the adsorption of DNP, Trp-P-1 or MeIQx. It also did not affect the adsorption of DNP to an insoluble, dietary-fibre preparation from wheat straw, which contains the hydrophobic component, lignin. The results are discussed in terms of hydrophobic interactions between carcinogens and insoluble dietary fibres. In vivo, it is likely that soluble dietary fibres reduce the adsorption of only highly hydrophobic carcinogens to some insoluble dietary fibres.