Feed pelleting improves digestibility and metabolic use of energy. As starch provides almost 70% of the dietary energy, the study aimed to test the effects of starch origin and pelleting on metabolism of pig. Six diets were obtained in a 3×2 factorial design differing by starch source (wheat, W; maize, M or barley, B) and presentation (flour, F or pelleted, P). The diets were named as the combination of botanical source and presentation form. Diets were distributed four times a day, providing 2.4 MJ ME/kg BW0.60 per day to 72 male pigs (mean body weight (BW): 52.2±4.7 kg) housed in pairs during one week in an open-circuit respiration chamber to measure the dynamics of gas exchanges and energy balance. Pigs were fitted with a catheter in the portal vein to study the dynamics of blood metabolites after a meal. At the end of the experimental period, pigs received two meals with indigestible markers 6 and 1 h before slaughter for quantifying marker recovery in each digestive tract segment. Marker passage in stomach 6-h after intake was higher with WP, MF and MP than with WF, BF and BP (87 vs 73%; P<0.05). Passage rate in stomach of marker given 1-h before slaughter were similar with WP, MF, MP and BF but higher than with WF and BP (34 vs 14 and 18%, respectively; P<0.05). Dry matter faecal digestibility was greater with WP and BP (+1.9 and +1.0%; P<0.05) compared to WF and BF. Preprandial portal insulinemia was higher with MP than MF (54 vs 9 μU/ml; P<0.05). From 150 to 180 min after the meal, insulinemia was greater with WP and BP than WF and BF (41, 29, 12.5 and 16.4 μU/ ml; P<0.05). Mean portal glycaemia was greater with BP than BF from 30 to 150 min after the meal (1.30 vs 1.47 g/l) and tended to be higher with WP compared to WF (1.44 vs 1.66 g/l). Pelleting improved digestive utilization and modified nutrient metabolic dynamics with W- and B- but not M-based diets. Varying responses to digestive and metabolic use between botanical sources have to be taken into account for feed formulation.