The objective of this study was to compare the clinical and bacteriological cure rates of cows with clinical mastitis following treatment with either tylosin base (5 g injected 3 times at 24-h intervals; n = 306) or penethamate hydriodide (5 g injected 3 times at 24-h intervals; n = 289). Duplicate milk samples were collected before treatment and again 14 +/- 3 and 21 +/- 3 d later for microbiological analysis. Only those quarters from which gram-positive mastitis pathogens were isolated before treatment were included in the analyses. Streptococcus uberis was the most prevalent isolate. The number of cows with clinical failure (i.e., retreated within 21 d of enrollment) did not differ between treatments (64 vs. 63, respectively). At the quarter level, there was no difference in the proportion of bacteriological cure between treatments (81.2 vs. 83.8% for penethamate hydriodide or tylosin, respectively). The proportions of clinical and bacteriological cure were influenced by age, herd, severity of mastitis, number of glands within the cow with clinical mastitis, bacterial species, and days postpartum at enrollment. There was no difference between treatment groups for SCC (4.46 vs. 4.44 +/- 0.08, mean +/- standard error of the difference in ln SCC for cows treated with penethamate hydriodide or tylosin, respectively) or production of milk solids (1.45 vs. 1.48 +/- 0.02 kg/d of milk fat + protein, for the penethamate hydriodide or tylosin treatment, respectively). Overall, there was no difference in the proportions of clinical failure (17.3 vs. 16.5% of cows treated with penethamate hydriodide or tylosin, respectively) or bacteriological cure (79.8 vs. 82.0% of cows treated with penethamate hydriodide or tylosin, respectively), or in SCC or milk production between dairy cows with clinical mastitis and those treated for clinical mastitis with 1 of 2 parenteral antibiotic therapies.