To determine prececal starch digestibility and estimate glucose uptake from the digestion of 4 forms of barley in the small intestine, 4 mature cecally fistulated geldings (449 ± 41 kg BW) fed a 62:38 (w/w) meadow hay:concentrate diet at 1.7 kg DM/100 kg BW were included in a 4 x 4 Latin square design experiment. During each period, horses received 80% DM of their concentrate as one of the 4 forms of a same batch of barley, whole grain, 2.5-mm ground, steam-flaked and pelleted. Hay was offered in 2 equal meals and concentrate in 2 unequal meals. The starch supply in the morning meal amounted 2.7 g starch/kg BW. At each period, mobile bag DM and starch disappearance was determined. Except for ground barley, each form of barley was 4-mm ground before being introduced in the bag. Nylon bags containing each substrate were intubated in the horse receiving the pelleted barley. Bags were collected in the cecum for 10 hours post-intubation. At each period, post-prandial glycemia was measured on blood samples collected on the 4 horses via an indwelling jugular catheter just before the concentrate morning meal and for 8 h. No hay in the morning meal was given the day of the measurements. Whole blood glucose was analyzed with a portable blood glucose meter. Mobile bag prececal DM disappearance and starch disappearance depended (P < 0.01) on barley form. Prececal starch disappearance of whole barley was the lowest but no difference (P > 0.05) was detected among the 3 processed grains. No significant effect of barley form was found whatever the glycemic parameters. No significant correlation was reported between glycemic parameters and the amount of prececal mobile bag disappeared starch calculated as the starch intake in the morning meal by the mobile bag starch disappearance. To conclude, the whole form of barley exhibited the lowest prececal mobile bag starch disappearance whereas, in relationship with large individual variations, no significant variation has been shown in glycemic parameters. Further investigations should be carried out to improve methods for estimating prececal starch digestion of processed cereals in the different digestive segments of horses.