The effects of using a water-soluble adjuvant or an emulsified oil-based adjuvant on the safety, antibody titer, and clinical responses of an Escherichia coli J5 bacterin were tested in an experimental infection trial. Fifty-one cows were assigned to 17 blocks of 3. Two cows within each block of 3 were vaccinated with a commercially prepared E. coli J5 bacterin containing either a water-soluble adjuvant or the same bacterin preparation emulsified in oil. One cow in each block was an unvaccinated control. Cows were immunized at drying off and 42 d later. The right or left front mammary quarter of each experimental cow was challenged by intramammary infusion of E. coli 727 between 14 and 35 DIM. Areas of inflammation at the primary injection site were greater 1, 2, and 3 d following primary vaccination for bacterin containing oil-in-water adjuvant compared with bacterin containing water-soluble adjuvant. Whey anti-E. coli J5 IgG titers were higher at calving for cows vaccinated with bacterin containing oil-in-water adjuvant than for cows either vaccinated with bacterin containing water-soluble adjuvant or unvaccinated controls. Serum x-E. coli J5 IgG titers were higher at calving for vaccinated cows than for unvaccinated controls. Peak bacterial counts in milk from challenged quarters were greater for unvaccinated controls than for cows vaccinated with bacterin containing water-in-oil adjuvant. Bacterial counts in milk from challenged quarters and clinical score both were greater in unvaccinated controls than cows vaccinated with bacterin containing water-in-oil adjuvant between 12 and 24 h postchallenge. Clinical responses were similar between unvaccinated controls and cows vaccinated with bacterin containing water-soluble adjuvant.