Selective use of antibiotic dry cow treatment can be implemented at the cow or quarter level, with the latter having the potential to further reduce antibiotic use. Our objective was to compare these 2 approaches in 6 herds in the United Kingdom in which environmental mastitis predominated. Eight hundred seven cows were enrolled and categorized as having a high cell count (n = 401) or low cell count (n = 406) in the last 3 mo of lactation and clinical mastitis history. All quarters of all enrolled cows received an internal teat sealant. Within each category, cows were randomly allocated to 1 of 3 groups; in one group antibiotic treatment was allocated at cow level (i.e., all 4 quarters received antibiotic), whereas in the 2 remaining groups antibiotic treatment was allocated at quarter level, based on California Mastitis Test (CMT) findings. Two different thresholds, score 1 and 2, were used to determine likely infection status. Quarter milk samples were collected at dry off and postcalving for bacteriological culture and somatic cell count (SCC). Cows were monitored for clinical mastitis from dry off until 100 d in milk. Cow level SCC and milk yield data were collated from farm records. Within each category, the 2 quarter level treatment groups were compared with cow level treatment at dry off. Leaving quarters untreated with intramammary antibiotic in cows in the high cell count group, with a CMT <2 or <1, reduced antibiotic use by 55% and 31%, respectively, and resulted in no difference in the odds of being infected with any pathogen postcalving, but was associated with a higher SCC at the first test day. Intramammary antibiotic treatment of quarters with a CMT ≥1 in cows in the low cell count category at dry off was not associated with any reduction in the odds of being infected with a major pathogen postcalving but was associated with a decrease in the odds of being infected with a minor mastitis pathogen postcalving. The use of antibiotics in quarters of cows categorized as low cell count at dry off, increased the proportion of quarters treated with antibiotic from 0% at cow level to 31% (CMT ≥ 1) and 12% (CMT ≥ 2) at quarter level, only resulting in a reduction in SCC of around 20,000 cells/mL at the first test day, if all quarters with CMT score ≥1 were treated with antibiotic. No differences in clinical mastitis incidence and milk yield in the first 100 d in milk were detected between any of the treatment groups. These study findings support selective quarter level dry off treatment only in cows with cow level SCC >200,000 cells/mL at dry off.