© 2016 Elsevier B.V. Highly fermentable fibre sources like sugar beet pulp (SBP) might be fed as an alternative to starch rich concentrate to horses. However, including soluble fibre from SBP in a meal of concentrate might delay gastric emptying and increase viscosity in the small intestine, altering or impairing glucose absorption, hence affect the metabolic responses measured in plasma and the hindgut. Four diets with different carbohydrate composition were investigated in a 4 × 4 Latin square design experiment in four periods using four caecum cannulated Norwegian coldblooded trotter horses. The diets were hay only (HAY), hay and molassed SBP (HAY + SBP), hay and pelleted barley (BAR), and hay, pelleted barley and molassed SBP (BAR + SBP). The amount of barley (2 g starch/kg body weight (BW)) fed in the test meals was similar for the BAR and BAR + SBP diets. Each diet was fed for 16 days followed by data collection, then each horse changed to a new diet in the next period. Caecum pH and short-chain fatty acid (SCFA) profiles and the metabolic responses measured in plasma at rest were followed for 9 h after the test meal. Caecum pH decreased more (P <0.001) after feeding BAR and BAR + SBP than after HAY and HAY + SBP. Diet also affected the proportions of acetate (P <0.001) and propionate (P <0.001) with less acetate and more propionate measured on the BAR and BAR + SBP diets than on the HAY and HAY + SBP diets, reflecting differences in the substrates fermented by the caecal microbes. Feeding barley resulted in a postprandial increase in plasma glucose (P = 0.030) and insulin (P <0.001) whereas these values remained almost stable when the fibre-based diets were fed. The total SCFA concentration in the caecum was greater (P <0.001) on the BAR and BAR + SBP diets than on the HAY and HAY + SBP diets. However, the plasma concentrations of SCFA were differently affected with greater (P <0.001) levels of SCFA (mainly acetate) in horses fed the HAY and HAY + SBP diets than in the BAR and BAR + SBP fed horses, reflecting a shift in the provision of carbohydrate derived nutrients from the small to the large intestine in response to the dietary carbohydrate composition. In conclusion, there was no effect of adding molassed SBP to a meal of barley compared to feeding barley alone, and fluctuations in plasma and caecal variables were more stable when feeding hay and molassed SBP than feeding barley.