In this study, the effectiveness of direct-fed microbials at reducing Escherichia coli O157 and Salmonella in beef cattle was evaluated. Steers (n = 240) received one of the following four treatment concentrations: control = lactose carrier only; low = 1 × 107 CFU per steer daily Lactobacillus acidophilus NP51; medium = 5 × 108 CFU per steer daily L. acidophilus NP51; and high = 1 × 109 CFU per steer daily L. acidophilus NP51. Low, medium, and high diets also included 1 × 109 CFU per steer Propionibacterium freudenreichii NP24. Feces were collected from each animal at allocation of treatment and found to have no variation (P = 0.54) between cohorts concerning E. coli O157 recovery. Feces and hide swabs were collected at harvest and analyzed for the presence of E. coli O157 by immunomagnetic separation and Salmonella by PCR. No significant dosing effects were detected for E. coli O157 recovery from feces at the medium dose or from hides at the medium and high doses. E. coli O157 was 74% (P < 0.01) and 69% (P < 0.01) less likely to be recovered in feces from animals receiving the high and low diets, respectively, compared with controls. Compared with controls, E. coli O157 was 74% (P = 0.05) less likely to be isolated on hides of cattle receiving the low dose. No significant dosing effects were detected for Salmonella recovery from feces at the medium and low doses or from hides at any doses. Compared with controls, Salmonella was 48% (P = 0.09) less likely to be shed in feces of cattle receiving the high dose. No obvious dose-response of L. acidophilus NP51 on recovery of E. coli O157 or Salmonella was detected in our study.