White-tailed deer (Odocoileus virginianus) in Michigan, USA, is one of the few species in the world considered a maintenance host for bovine tuberculosis (TB). Prevalence of TB over the past decade has remained steady in Deer Management Unit (DMU) 452 and new control strategies are needed to further reduce prevalence or eradicate the disease. There is little public support for large-scale culling of deer in Michigan, and vaccination is being considered as a potential management tool. Our goal was to evaluate the use of vaccination and/or targeted removal via live-trapping as TB control measures and how changes in vaccine parameters affect prevalence reduction. We used a spatially explicit stochastic individual-based model of TB dynamics in white-tailed deer to evaluate 39 management scenarios for their ability to 1) reach a 95% probability of eradication, and 2) lower TB prevalence over 30 years. We found continuation of current management would not result in a detectable prevalence reduction, nor would treatment of a “hot-spot” of infection within DMU 452. Application frequency played a larger role in prevalence reduction than vaccine efficacy, guaranteed immunity period, or half-life. Vaccination (alone or with targeted removal) applied annually throughout DMU 452 was necessary for a detectable reduction in prevalence. Projected costs for application over DMU 452 would exceed US$1.5 million annually. Vaccination will be costly and require long-term commitment, but the economic losses associated with allowing TB to persist in Michigan could outweigh vaccination costs. © 2012 The Wildlife Society.