Wildlife managers use models to aid in predicting high risk areas for human and black bear (Ursus americanus) interactions (HBI). These tools help managers implement management strategies to minimize HBI. Over 3,000 incidents of HBI were compiled from management reports at Great Smoky Mountains National Park (GSMNP) during 1998-2011, a park with 9-10.2 million visitors per year and a black bear population of about 1,600 bears. We used data from bear management reports along with annual visitor use, mast and bear abundance data to develop a series of generalized linear models to assess the spatial and temporal factors affecting HBI. Although HBI occurred throughout the GSMNP, 50% of all HBI occurred in five areas. The best predictor variables of HBI across four subsets of models included interaction between mast production and number of park visitors, month, vegetation cover, visitor activity, and bear abundance. Although there was not a clear relationship between visitor use and mast abundance, the number of park visitors was always relatively high and HBI increased substantially in poor mast years. HBI was more frequent during summer months when park visitation rates and more people and food were present overnight in frontcountry and backcountry camping areas. Over 43% of HBI in hemlock forests were serious. Bear abundance data were not a strong predictor of HBI, and bear bait stations may not provide a sensitive index to bear abundance. GSMNP uses different strategies for managing HBI to protect visitors and bears. In 1991, bear proof waste disposal containers and food storage devices were placed in camping and picnic areas. In combination with aversive conditioning, HBI decreased in some areas of the park. We recommend that proactive bear management programs including education, enforcement of park regulations, and aggressive aversion conditioning of bears be implemented at the identified HBI high risk areas to provide a safer environment for both people and bears in GSMNP.