The objective of this study was to verify the effect of mean particle size (mPS) on both gas production and in vitro rumen starch degradability (IVSD) of corn and barley meals (Cm and Bm, respectively). Batches of the same Cm or Bm were separately processed through 2 different mills (i.e., a cutter mill or a rotor speed mill) equipped with or without different screens to achieve different mPS for each tested meal. Samples were analyzed accordingly to a completely randomized design and the main tested effect of model was mPS (n = 11, from 0.46 to 3.50 mm mPS for Cm or n = 10, from 0.11 to 2.98 mm mPS for Bm). For both in vitro assays, the rumen inocula were collected from 2 rumen-fistulated Holstein lactating dairy cows fed a total mixed ration with 16.2% crude protein, 28.5% starch, and 35.0% neutral detergent fiber on a dry matter basis. To fit gas production data, 1-pool exponential model and 1-pool or 2-pool Gompertz models were adopted. The rate of gas production decreased and lag increased by increasing mPS of both Cm and Bm, irrespective of adopted 1-pool models. When the 2-pool Gompertz model was used to fit gas production data, a shift of particles from fast to slow fermentable pools was measured by increasing mPS. In particular, the ratio between fast and slow final volumes ranged from 0.90 at 0.11 mm mPS to 0.10 at 2.98 mm mPS for Bm. For Cm, the ratio between fast and slow final volumes decreased quadratically by increasing mPS, with the highest value (i.e., 0.58) measured at the lowest tested mPS. Values lower than 0.10 were measured for mPS greater than 1.93 mm for Cm. Concerning IVSD data, linear decreases in rate of starch degradation equal to -0.049 or -0.092 h-1 for each 1-mm increase in mPS were achieved for Cm and Bm, respectively. The 7-h IVSD decreased by 6.3 or 6.5% starch for each 1-mm increase in mPS of Cm or Bm, respectively. Present findings supported the hypothesis that different particle sizes within the same starch source represent an important factor influencing both fermentation kinetic parameters and IVSD.