Diagnosis techniques and treatments for Escherichia coli mastitis in the dairy bovine were evaluated in two experiments. The first experiment evaluated eosin methylene blue agar as a method of distinguishing E.coli from other gram-negative mastitis pathogens. Escherichia coli will usually produce a green metallic sheen on eosin methylene blue agar. One hundred and twenty-nine milk samples or gram-negative isolates from milk samples were used to compare eosin methylene blue agar to a commercial biochemical test strip (the accepted standard). There was an intermethod agreement of 96.9% and a k-value of 93.7% indicating excellent agreement beyond chance between test methods. Eosin methylene blue agar is a reliable method for differentiation of E. coli from other gram-negative mastitis pathogens. The second experiment evaluated the efficacy of frequent milk-out as a treatment for E. coli mastitis. Sixteen Holstein dairy cows were divided into 2 blocks and randomly assigned to 1 of 4 treatment groups: 1) non-infected, not frequently milked-out, i.e. not treated (NI-NT), 2) experimentally infected with E. coli, not treated (EC-NT), 3) non-infected, frequently milked-out (NI-FMO), and 4) experimentally infected with E. coli, frequently milked-out (EC-FMO). Hours to bacterial, clinical and systemic cure were not different between the EC-NT and EC-FMO treatment groups. Serum a-lactalbumin concentrations were evaluated between treatment groups as a measure of udder health. Serum a-lactalbumin concentrations were higher in cows in the EC-NT treatment group than cows in the NI-NT, NI-FMO and EC-FMO treatment groups at 12 hours post-experimental challenge. Serum a-lactalbumin concentrations were higher in cows in the NI-FMO treatment group than in cows in the NI-NT, EC-NT and EC-FMO treatment groups at 36 hours post-experimental challenge. Results from this study do not support frequent milk-out as a treatment for E. coli mastitis. / Master of Science