Steers fitted with simple rumen and abomasal cannulas were given isoenergetic diets of approximately equal amounts of untreated (UT) barley straw and concentrates (flaked maize +tapioca) alone (BS) or with urea (BSU) or fishmeal (BSF). Similar diets were also given in which the barley straw had been treated (AT) with NaOH (BSA, BSAU and BSAF respectively). The diets were given in a 6 X 6 Latin square design. Feed components and abomasal digesta samples were analysed for neutral (NDF) and acid (ADF) detergent fibres and for monosaccharide constituents of structural polysaccharides. Hemicellulose contents were estimated as the sum of xylose + arabinose (X + A) and by the difference between ash-free NDF and ash-free ADF (NDF-ADF). Cellulose was estimated as @-linked glucose (C) and by the difference between ash-free ADF and lignin (ADF-L).lo3 Ruthenium and PEG were given as flow markers and flows (g24h-') at the abomasum of carbohydrate components estimated in these ways were calculated. Approximately 98% (by wt.) of the cellulose (C) found in original feed and digesta samples was recovered in both NDF and ADF. Recoveries of hemicellulose (X + A) in NDF from UT straw, AT straw and abomasal digesta were approximately 92, 48 and 50%, respectively. The ADF fraction of feeds and digesta contained 3-6 and 10-17% of the nitrogen and xylose, respectively, present in the original samples. Mouth to abomasum digestibilities of hemicellulose (NDF-ADF) for diets BS, BSU, BSF, BSA, BSAU and BSAF were 39, 62, 67, 29, 61 and 76%, respectively. Corresponding values for cellulose (ADF-L) were 37, 34, 50, 45, 48, and 63%, respectively. The use of NDF-ADF and ADF-L as measures of hemicellulose and cellulose contents, respectively, of feeds and digesta, and the digestibility of these carbohydrate fractions between mouth and abomasum of steers are discussed.