Bovine tuberculosis (bTB) caused by Mycobacterium bovis infection in Michigan white‐tailed deer (Odocoileus virginianus) has proven resistant to current management practices. The Michigan Department of Natural Resources (MDNR) is faced with managing a protracted bTB outbreak with shrinking economic resources, its initial control strategies approaching, or having reached, the limits of their effectiveness. Planning tools are needed to project the outbreak's future course and forecast the likely outcomes of proposed controls. We describe development of a spatially explicit, individual‐based stochastic simulation model of bTB in Michigan white‐tailed deer. We sought to 1) characterize whether eradication of bTB is possible by increasing hunter harvest or via vaccination, and how long it is likely to take to achieve eradication; 2) characterize the effect of concurrent deer baiting; and 3) assess the effect of baiting on the probability of bTB establishment in uninfected areas. Simulations indicated that current MDNR management strategies are unlikely to eradicate bTB from the core outbreak area's deer population within the next 30 years. A 50–100% increase (over current rates) of both antlered and antlerless deer harvest was required to achieve eradication if baiting was occurring, compared to only a 50% increase in harvest required if baiting was eliminated. Vaccination strategies required frequent application and high exposure rates (>90%) to achieve eradication, which baiting delayed. Simulations indicated that if bTB was eradicated from the core outbreak area, a single infected deer introduced into the area would be 8 times more likely to re‐establish bTB if baiting was occurring. The ability to forecast likely outcomes of disease management can be critical for wildlife managers to assess whether specific strategies are likely to be successful. Because current policy appears unlikely to achieve the stated goal of eradicating bTB from Michigan in the foreseeable future, reorienting the bTB program from eradication to controlling geographic spread and transmission to cattle may be more realistic goals. Spatial models such as ours are ideally suited to investigating spatial heterogeneity of disease transmission, and how transmission is influenced by aggregating factors such as baiting or supplemental feeding. © 2014 The Wildlife Society.