The aim of this study was to investigate the cattle—exposure factors associated with the risk of a bovine animal reacting to a bovine tuberculosis (bTB) skin test at a whole herd test. There were 148 study farms enrolled. These were located in six counties of the south west of England in an area considered endemic for bovine tuberculosis (bTB): 24% were restocked after foot and mouth disease (FMD) in 2001; all farms were located within the Randomised Badger Culling Trial (RBCT) area. Data on cattle on these farms were sourced from the bTB Vetnet database from 1996 to 2004 and from the British Cattle Movement Scheme database. Individual animal records were created that included data on whether or not an animal became a reactor at a full-herd bTB test between 1 June 2001 and 19 August 2004, their prior exposure to cattle with bTB (defined by presence at a bTB test where at least one reactor was detected), whether the animal was homebred, the farm history of bTB and the farm restocking status. Data from 144 farms were used, 4 farms had no data.Cattle were more likely to react to the bTB skin test when they had been present at a previous bTB herd test (or tests) where other cattle had reacted to the skin test. This positively correlated with age and the number of bTB tests an animal had had. Cattle on restocked farms were less likely to react to the skin test compared with cattle on continuously stocked farms. These results highlight the likely importance of exposure to infected cattle at a previous test as a source of infection to cattle that subsequently became reactors and suggest that there was a lower risk of exposure to bTB to cattle in newly formed herds.