This experiment evaluated dry matter intake, ruminal in situ nutrient disappearance, total-tract apparent nutrient digestibility, microbiome, and ruminal fermentation parameters of crossbred beef cows offered forage-based diets and a Bacillus-based direct-fed microbial (DFM). Fifteen rumen-cannulated Angus × Hereford cows were used in a crossover design. Treatments consisted of forage-based diets plus 500 g of protein-mineral supplement mixed with 1) 3 g of a Bacillus-based DFM containing Bacillus licheniformis and B. subtilis (BAC; n = 15; 2.2 × 109 CFU of the mixture/g; Bovacillus, Novonesis, Lyngby, Denmark), or 2) without BAC (CON; n = 15). Each experimental period lasted 33 days with 30 days of wash-out between periods. Cows were fed daily with chopped grass-mixed hay [Idaho fescue (Festuca idahoensis) and Bluebunch wheatgrass (Pseudoroegneria spicata)]. From day 17 to 26, 5 g of titanium dioxide (TiO2) was dosed twice daily, at 12-h intervals, as an external marker of fecal output. From day 22 to 26, fecal spot samples were collected at 12-h intervals. On day 27, ruminal fluid was collected prior to feeding (0 h) and at 4, 8, 12, 16, and 20 h post-feeding to evaluate ruminal pH, volatile fatty acids (VFA) and NH3-N. From day 28 to 32, 4 g of grass-mixed hay was placed in Dacron bags and introduced through the cannulas for 0, 12, 24, 48, 72, and 96 h. A treatment effect was observed (P = 0.01) for DMI, which was greater for cows supplemented with BAC vs. CON. Cows supplemented with BAC tended (P > 0.06) to have greater DM and NDF disappearance vs. CON. A treatment effect was observed (P ≤ 0.02) for total-tract apparent DM and CP digestibility, which were greater for BAC vs. CON cows. Also, cows supplemented with BAC tended (P ≥ 0.06) to have a greater total-tract NDF and ADF digestibility vs. CON cohorts. Therefore, cows supplemented with BAC had greater (P < 0.01) amounts of DM, CP, NDF, and ADF digested vs. CON cows. A treatment × time effect was observed (P = 0.02) for ruminal pH, which increased at 16 and 20 h of collection for BAC vs. CON cows. Treatment effects were not (P ≥ 0.18) observed for VFA, N-NH3, and plasma concentrations of glucose, whereas plasma concentration of urea tended (P = 0.06) to be reduced for cows supplemented with BAC vs. CON. Hence, supplementation with a Bacillus-based DFM increased forage intake, the digestibility of DM and CP, and the amount of nutrients digested by rumen-cannulated cows receiving a forage-based diet.