In October 1985, mycoplasmas were isolated from bulk tank milk samples in a large Florida dairy (greater than 1,400 lactating cows). At that time, measures to isolate and control the spread of infection were instituted. In an initial screening test, Mycoplasma bovis was isolated from 21 of 153 milking string samples (milk from all quarters of 10 cows/string). Composite quarter milk samples from all quarters of every individual lactating cow in the herd were obtained for culture in November 1985 and December 1985. In October, 88 of 1,535 (5.7%) cows were identified as Mycoplasma-positive. An additional 31 Mycoplasma-infected cows were identified in December. The dairy elected to maintain the infected cows in a separate Mycoplasma-positive subherd, which would be milked at the end of each milking session. Seven additional Mycoplasma-positive cows were identified at initiation of lactation. All newly identified infected cows were transferred to the Mycoplasma-positive subherd. After segregation of Mycoplasma-positive cows, bulk tank milk samples obtained routinely from the main herd remained culture negative throughout the study. From February 1986 to October 1986, quarter milk samples were obtained monthly from cows in the Mycoplasma-positive subherd. Any cow that developed clinical mastitis or substantial decrease in milk production was, at the discretion of the herdsman, culled. Of the 126 cows in the subherd, 22 (17.5%) were culled for mastitis, 35 (27.8%) were culled for low production, and 9 (7.1%) were culled for other reasons. Of the remaining 60 cows, 16 (12.7% of the 126 cows) were Mycoplasma-positive on the basis of results from one or more samples obtained after February 1986.(ABSTRACT TRUNCATED AT 250 WORDS)