Two experiments evaluated the effect of a direct-fed microbial (Bovamine, Chr. Hansen, Hørsholm, Denmark) on performance and digestion of dairy cows. In Experiment 1 (Exp1), 30 multiparous cows (40–160 DIM) were assigned to one of 2 treatments fed for 10 weeks, DFM (3 × 10⁹ cfu/d Bovamine) or CON (control, no Bovamine). In Experiment 2 (Exp2), 6 ruminally cannulated cows (123 ± 129 DIM) were assigned to a crossover design with 6-wk periods with the same treatments as Exp1 except that cows were fed a 23.8% starch diet during wk 1–5 of each period and then abruptly switched to a 31.1% starch diet for wk 6. For both experiments, intake and milk yield were measured daily and milk samples were collected weekly. In Exp1, fecal grab samples were collected every 6 h on d 7 of wk 1, 2, 4, 6, 8, and 10, fecal consistency was scored, and fecal starch measured in composited daily samples. Composites from a subset of 7 cows per treatment were used to measure apparent total-tract nutrient digestibility. In Exp2, rumen pH was continuously recorded during wk 5–6. Rumen in situ digestibility was measured on wk5d7, wk6d1, and wk6d7. On those dates, rumen fluid and feces were collected every 6 h for rumen VFA, fecal pH, and fecal starch (composited by cow within day). In Exp1, treatment did not affect intake, milk yield or composition, fecal score or fecal starch. DFM tended to increase starch digestibility compared with CON (98.74 vs. 98.46%, P = 0.051), but digestibility of other nutrients was unaffected. In Exp2, intake, milk yield, and milk composition were unaffected when evaluated over the entire study. However, during the abrupt switch to the high starch diet, milk fat yield was increased for DFM compared with CON (1.39 vs. 1.28 kg/d, P = 0.002) and milk fat tended to increase (3.59 vs. 3.42%, P = 0.09). Treatment did not affect rumen pH, rumen VFA, or in situ digestibility. Contrary to Exp1, DFM increased fecal starch compared with CON during the high starch feeding (2.49 vs. 2.03%, P = 0.02). Overall, Bovamine modestly improved digestion as demonstrated by increased starch digestibility in Exp1 and increased milk fat during the high starch challenge in Exp2.