Implementation of protocols for prudent use of antimicrobials on dairies results in economic returns, sustainability of products, and addresses public health concerns such as resistance and residues. Two main areas of application are intramammary (IMM) treatment of clinical mastitis (CM) and dry-cow therapy, which account for 80% of antibiotic use on dairies. The objectives herein were to 1) evaluate if the antibiotic hetacillin fell within a margin of efficacy or “noninferiority” to the reference treatment, a critically important (to human medicine) third-generation cephalosporin, ceftiofur; 2) assess a protocol that selectively treats CM cases based on daily culture results; 3) determine if similar protocols can be effectively executed on moderate-sized dairies without daily access to a professional laboratory; 4) determine if the metagenomic profiles of CM cases highlight a missed opportunity for non-treated cases; and 5) evaluate a culture-independent selective dry-cow therapy (SDCT) protocol. Noninferiority of hetacillin was concluded for bacteriological and clinical cures, as well as for many secondary outcomes. Administration of hetacillin over ceftiofur decreased treatment pen time with no statistical differences in bacteriological, pathogen, or clinical cures; milk yield and linear score (LS) postevent; nor additional risk of culling or quarter-level mastitis after the event. Use of a selective pathogen-based protocol decreased milk withholding time with no statistical differences in days to clinical cure, milk yield, and LS postevent; nor additional odds of culling in the months following. When similar strategies were applied on 8 moderate-sized dairies using on-farm or 5 day/week diagnostics, irregularity of data recording was drastically improved and no negative changes were reported. Reductions in antimicrobial use ranged from 6.8% to 92.5% and economic savings amounted to $2,200 to $4,600/dairy/yr. 16S rRNA analysis validated that despite dysbacteriosis in the mammary gland, there was no pathogen profile that was associated with negative or positive health and production outcomes; missed opportunities were not apparent. Finally, application of a data-driven, SDCT protocol resulted in no statistical differences in new infection risk, milk production and LS, or additional odds of culling or mastitis over the periparturient period between selectively or blanket-treated low-risk cows. Using SDCT or pathogen-based protocols for the treatment and prevention of CM can reduce antibiotic use by up to 2/3.