Wildlife managers often rely on permanent or temporary area closures to reduce the impact of human presence on sensitive species. In 1982, Yellowstone National Park created a program to protect threatened grizzly bears (Ursus arctos) from human disturbance. The bear management area (BMA) program created areas of the park where human access was restricted. The program was designed to allow unhindered foraging opportunities for bears, decrease the risk of habituation, and provide safety for backcountry users. The objective of our study was to evaluate human-bear interaction in BMAs and determine if they were effective. We used human and grizzly bear global positioning system location data to study 6 of 16 BMAs from 2007 to 2009. We contrasted data when BMAs were unrestricted (open human access) and restricted (limited human access). We used location data collected when BMAs were unrestricted to delineate a human recreation area (HRA) and determined a daily human active and inactive period. We applied the HRA and daily activity times to bear location data and evaluated how bear movement behavior changed when people were present and absent. We found that grizzly bears were twice as likely to be within the HRA when BMAs were restricted. We also found that grizzly bears were more than twice as likely to be within the HRA when BMAs were unrestricted, but people were inactive. Our results suggest that human presence can displace grizzly bears if people are allowed unrestricted access to the 6 BMAs in our study. Our study provides evidence for the utility of management closures designed to protect a threatened species in a well-visited park. Our approach can be reapplied by managers interested in balancing wildlife conservation and human recreation. © 2013 The Wildlife Society.