The influence of barley brans on the incidence and burden of intestinal tumours in rats induced by 1,2-dimethylhydrazine (DMH) was studied in a 7 month feeding experiment. The basic diet was American Institute of Nutrition (AIN) 76 modified by adjustment to 20% fat and 40% starch; brans were added so as to supply 5% dietary fibre. The barley brans studied were commercial barley bran (BB1; 13.0% dietary fibre) from the aleurone/subaleurone layer, outerlayer barley bran (BB2) including the germ (25.5% dietary fibre) and spent barley grain bran (SBG; a by-product of the brewery and including the hull; 47.7% dietary fibre). They were compared with wheat bran (WB; 44% dietary fibre) and cellulose (or control; 98% dietary fibre). Commercial barley bran and wheat bran were most effective in reducing tumour incidence and burden. The incidence of tumours fell significantly from 70% (BB2) and 50% (SBG) to 10% (BB1) and 20% (WB) and tumour burden and tumour mass index (TMI) were also reduced by similar orders of magnitude. There were significantly higher short chain fatty acid (SCFA) concentrations in WB- and BB1-fed rat faecal pellets relative to cellulose- and BB2-fed rat faeces; butyrate, in particular, was affected. Regression analysis of butyrate against tumour incidence showed a trend (r = 0.898: P = 0.055), but the concentration of butyrate alone could not account for the reduction in tumour incidence observed. In a second experiment, when two brans (BBI and SBG) were introduced after DMH dosing, there were higher incidences and burdens of tumours, indicating that protection by such brans was not as effective under these circumstances. Commercially available barley bran and wheat bran appear to significantly reduce tumour incidence and burden in this model relative to other brans, influencing both the initiatory as well as promotional stages of chemically induced carcinogenesis.