Identifying the presence of bovine tuberculosis (TB; Mycobacterium bovis) in wildlife is crucial in guiding management aimed at eradicating the disease from New Zealand. Unfortunately, surveys of the principal wildlife host, the introduced brushtail possum (Trichosurus vulpecula), require large samples (> 95% of the population) before they can provide reasonable confidence that the disease is absent. In this study, we tested the feasibility of using a more wide-ranging species, feral pig (Sus scrofa), as an alternative sentinel capable of indicating TB presence. In January 2000, 17 pigs in four groups were released into a forested area with a low density of possums in which TB was known to be present. The pigs were radiotracked at 2 wk intervals from February to October 2000, and some of them were killed and necropsied at various intervals after release. Of the 15 pigs successfully recovered and necropsied, one killed 2 mo after release had no gross lesions typical of TB, and the only other pig killed at that time had greatly enlarged mandibular lymph nodes. The remainder were killed at longer intervals after release and all had gross lesions typical of TB. Mycobacterium bovis was isolated from all 15 pigs by mycobacterial culture. Home range sizes of pigs varied widely and increased with the length of time the pigs were in the forest, with minimum convex polygon range-size estimates averaging 10.7 km2 (range 4.7-20.3 km2) for the pigs killed after 6 mo. A 6 km radius around the kill site of each pig would have encompassed 95% of all of their previous locations at which they could have become infected. However, one pig shifted 35 km, highlighting the main limitation of using unmarked feral pigs as sentinels. This trial indicates use of resident and/or released free-ranging pigs is a feasible alternative to direct prevalence surveys of possums for detecting TB presence.