The objective of the trial was to quantify the potential of a starch binding agent (BioProtect) to reduce in vitro rumen starch degradation of corn and sorghum particles varying in size. Corn and sorghum grain samples were ground through 1 and 2 mm sieves using a Wiley mill and subsequently sieved to obtain the following sizes: <250, 250–500, 500–1180 and 1180–2000 µm. All fractions were separately analyzed for starch. Samples were treated 24 h before fermentation by spraying with BioProtect according to product guidelines. Both, treated and untreated corn and sorghum samples were fermented in vitro for 0, 3, 6, 9, 12, 24 and 48 h to quantify starch digestibility (Sd). Rates of digestion (kd) were calculated using a first order decay model and 48 h fermentation residuals were used to estimate indigestible starch. Data were analyzed according to a randomized complete block design with a factorial arrangement of treatments. The main effects tested were grain, particle size, product, time (for Sd only) and their interactions. Fermentation run (n = 3) was considered a random effect. As particle size decreased, starch increased from 701 to 821 g/kg and from 730 to 810 g/kg for corn and sorghum, respectively. For both grain types, Sd and kd increased linearly with decreased particle size (P < 0.01). Sd (kd) increased from 41 (0.10) to 58% (0.26 h⁻¹) and 30 (0.11) to 53% (0.24 h⁻¹) for corn and sorghum, respectively. BioProtect was effective (P < 0.001) in decreasing starch digestibility for both grains. The product was more effective with smaller particle size, by reducing Sd 17%-units for the smallest particles vs. 7%-units for the largest particles. A time interaction was observed (P < 0.0001), showing that the highest impact of BioProtect occurred after 12 h of fermentation for both grains. The starch binding agent resulted in an effective decrease of in vitro starch digestion, but results would be affected by particle size and fermentation time.