This study investigated the effects of bacterial direct fed microbials (DFM) on ruminal fermentation and microbial characteristics, methane (CH4) emission, diet digestibility, and milk fatty acid (FA) composition in dairy cows fed diets formulated to induce different ruminal volatile fatty acid (VFA) profiles. Eight ruminally cannulated dairy cows were divided into 2 groups based on parity, days in milk, milk production, and body weight. Cows in each group were fed either a high-starch (38%, HS) or a low-starch (2%, LS) diet in a 55:45 forage-to-concentrate ratio on a dry matter (DM) basis. For each diet, cows were randomly assigned to 1 of 4 treatments in a Latin square design of (1) control (CON); (2) Propionibacterium P63 (P63); (3) P63 plus Lactobacillus plantarum 115 (P63+Lp); (4) P63 plus Lactobacillus rhamnosus 32(P63+Lr). Strains of DFM were administered at 10(10) cfu/d. Methane emission (using the sulfur hexafluoride tracer technique), total-tract digestibility, dry matter intake, and milk production and composition were quantified in wk 3. Ruminal fermentation and microbial characteristics were measured in wk 4. Data were analyzed using the mixed procedure of SAS (SAS Institute Inc., Cary, NC). The 2 diets induced different ruminal VFA profiles, with a greater proportion of propionate at the expense of acetate and butyrate for the HS diet. Greater concentrations of total bacteria and selected bacterial species of methanogenic Archaea were reported for the HS diet, whereas the protozoa concentration in HS decreased. For both diets, bacterial DFM supplementation raised ruminal pH (+0.18 pH units, on average) compared with CON. Irrespective of diet, P63+Lp and P63+Lr increased ruminal cellulase activity (3.8-fold, on average) compared with CON, but this effect was not associated with variations in ruminal microbial numbers. Irrespective of diet, no effect of bacterial DFM on ruminal VFA was observed. For the LS diet, supplementing cows with P63+Lr tended to decrease CH4 emission (26.5%, on average, when expressed per kilogram of milk or 4% fat-corrected milk). Only P63 supplementation to cows fed the HS diet affected the concentration of some milk FA, such as cis isomers of 18:1 and intermediates of ruminal biohydrogenation of polyunsaturated FA. Overall, bacterial DFM could be useful to stabilize ruminal pH. Their effects on CH4 production mitigation and milk FA profile depended on DFM strain and diet and should be confirmed under a greater variation of dietary conditions.