The digestion of polysaccharides from the wheat cultivars Caphorn and Isengrain was investigated, and the efficiency of an enzyme preparation was tested using the TNO gastrointestinal model (TIM-1). The apparent digestibility (AD) of carbohydrates was determined based on the measurement of organic matter (OM), total monosaccharides, reducing ends (RE) and end products (EP: glucose, maltose and xylobiose). The AD of the OM from Caphorn and Isengrain measured using caecectomised cockerels did not differ from that measured using TIM-1: 72·0 (sd 2·6) v. 70·6 (sd 0·6) % for Caphorn (P = 0·580) and 73·0 (sd 2·3) v. 71·1 (sd 1·9) % for Isengrain (P = 0·252). After the 6 h TIM-1 digestion, 41·4–58·9 % of the OM, RE and EP were recovered from the jejunal compartment and 18·3–27·1 % from the ileal compartment, while ileal deliveries and digestive residues constituted the remainder. A commercial enzyme cocktail tested at 0·2 μl/g of wheat improved TIM-1 digestibility of Caphorn and Isengrain polysaccharides: 3·9 % (P = 0·0203) and 3·4 % (P = 0·0058) based on the OM; 9·7 % (P < 0·0001) and 3·1 % (P = 0·031) based on the total glucose; 47·2 % (P < 0·0001) and 14·2 % (P = 0·0004) based on the RE, respectively. The enzyme cocktail improved the release of the EP for Caphorn (3·8 %, P = 0·008) but not for Isengrain ( − 0·8 %, P = 0·561). The higher efficiency of the enzyme supplementation on the digestion of Caphorn polysaccharides compared with Isengrain seems to be linked to the higher soluble carbohydrate contents and/or less ramified arabinoxylan of Caphorn.