The purpose was to evaluate 2 intramammary treatments for mild-to-moderate cases of clinical mastitis in a noninferiority comparison. Noninferiority trials are intended to show whether a given treatment, hetacillin potassium, has at least comparable efficacy as the reference treatment, ceftiofur hydrochloride. Treatments can be deemed inferior to the reference treatment by an amount less than the margin of noninferiority, or inconclusive if the confidence interval crosses the margin of noninferiority. Cows with clinical mastitis from 6 farms were considered for enrollment. Using a randomized design, cows with mild or moderate mastitis in 1 quarter were assigned to on-label treatment with either ceftiofur or hetacillin. A total of 596 cows met the criteria needed for continued enrollment. Treatment distribution resulted in 309 cows in the ceftiofur group and 287 cows in the hetacillin group. Mixed regression analysis was performed for the following outcomes: bacteriological cure, pathogen cure, clinical cure, postevent milk production and linear score, and survival to d 30 and 60. Cox proportional hazards analysis was used to describe treatment effect on survival and mastitis risks. Bacteriological cure, defined as absence of causative organism in samples retrieved at d 14 and 21 postmastitis, was similar between groups. No significant statistical differences were found in cure risk, and noninferiority of hetacillin relative to ceftiofur for bacteriological cure was conclusive (hetacillin=67%, ceftiofur=72%). Absence of a pathogen on both follow-up samples designated a cow as a pathogen cure. Pathogen cure was similar between treatment groups and noninferiority of hetacillin relative to ceftiofur was shown (hetacillin=35%, ceftiofur=32%). Clinical cure (hetacillin=68%, ceftiofur=64%), postevent milk production (hetacillin=37.0kg, ceftiofur=38.2kg), and linear scores (hetacillin=3.4, ceftiofur=3.1) were also not statistically different between treatment groups. Noninferiority of hetacillin relative to ceftiofur was shown for survival to d 30 and survival to d 60, whereas hetacillin was more likely to have a clinical cure than ceftiofur by d 4. No differences were seen between groups when Cox proportional hazards were performed, neither for exit from the herd in the 60 d following the event nor in the risk for a subsequent mastitis event. These findings can be used to develop farm-specific protocols for clinical mastitis treatment.