The objective of the study was to monitor the effect of two therapy regimens on experimental Escherichia coli mastitis. Single udder quarters of 12 cows that were at least 30 d postpartum were inoculated with 1500 cfu of E. coli. The inoculation was repeated in the contralateral quarter after a 3- to 4-wk interval. Initially, half of the cows were treated with antimicrobials, and the remaining half were left untreated. At the second inoculation, the cows that were originally treated were not treated, and vice versa. Therapy began 12 h after inoculation and consisted of parenteral trimethoprimsulfadiazine (6 cows) or intramammary colistin sulfate (6 cows). Clinical signs, daily milk yield, bacterial count, and endotoxin content of the milk were recorded. Milk SCC, NAGase activity, and trypsin inhibitor capacity were also monitored. The response to bacterial challenge varied greatly among cows. Bacteria were eliminated from the quarters within 7 d in all but 1 cow. Treatment did not significantly affect the elimination rate of bacteria or any of the measured parameters. Significant positive correlations existed among milk bacterial counts, endotoxin concentrations, and clinical signs at the acute stage of the infection. Based on these findings, antimicrobial therapy of E coli mastitis during lactation apparently is no more beneficial than no treatment.