In vitro (IV) starch digestion (IVSD) is used to evaluate starch sources for dairy rations. The study objective was to determine the efficacy of IVSD and IV gas production (IVGP) in estimating fractional rates of starch digestion (kds). Three high-moisture corn samples with high, medium or low IVSD and 2 dry corn samples with high and low IVSD were selected from 20 random commercial corn grains. Samples were oven-dried and ground using a Wiley mill (4-mm screen). Blended ruminal fluid from 3 steers fed a 30% starch diet was used to measure IVSD at 3, 8, and 24 h and IVGP from 0 to 24 h using an Ankom RF system. Duplicate samples were measured using the same inoculum and media in 3 IV runs. Across all samples and times, IVGP was highly correlated with grams of digested starch: IVGP (mL/g) = 388 × g digested starch; R² = 0.975, SE = 8.0; P < 0.0001. As a reference, MIR_kd for starch was calculated using logarithmic transformation (log-trans) and a constant lag with IVSD(8h). Reference MIR_kd were different among grains and IV runs. Both kds and lag can be calculated using log-trans with IVSD(3,8h) giving the relationship: MIR_kd(8h) = 0.014 + 0.8693 × kds(3,8h); R² = 0.991, SE = 0.0039, P < 0.001). Average lag(3, 8h) was 2.02 h and did not vary among grains, but varied between IV runs (P < 0.009). Solver (MS Excel add-in) was used to estimate lag, rate of gas production (kgp), and kds for IVSD(3,8,24h): MIR_kd(8h) = −0.0197 + 1.0348 × kds(3–24h); R² = 0.969, SE = 0.0071, P < 0.001. As with MIR_kd(8h), kds(3–24h) varied among runs (P < 0.001) and samples (P < 0.001). Variation between IV runs was less for IVGP compared with IVSD. When separate intercepts were allowed for each run, the prediction was: MIR_kd(8h) = IVrun + 0.9805 × kgp; R² = 0.861, SE = 0.0155, P < 0.001. High correlations between kds calculated from IVSD at 3, 8 and 24h and the small variation among grains in IVSD at 3 and 24 h suggest these times add little when estimating kds. The high correlation between IVGP and g of digested starch during fermentation, and the relationship between kds and kgp suggest that IVGP will be useful in defining the kinetics of carbohydrate digestion in grain.