Large carnivores are among the most threatened species in the world because of their natural low densities and need for expansive habitats. The brown bear (Ursus arctos) is the largest carnivore in the southwestern Asia, and faces threats in much of its range from conflict with humans over shared resources and shrinkage of habitat. In this study, we surveyed for brown bear sign and scat during spring–autumn from April 2013 to November 2015 in 24 randomly selected, 25-km² grid cells, and developed a model of potential brown bear occurrence in one of its globally southernmost distribution ranges in Iran. To better understand its conservation needs and management priorities at the landscape scale, we used a combination of field surveys to develop a Maximum Entropy (Maxent) model. The model was developed using 10 environmental and anthropogenic predictors. Potential brown bear occurrence was strongly influenced by availability of water resources (54.1%) as the most important variable; and distance to roads (16.1%), aspect (7.6%), and vegetation types (5.9%) were the other important factors. The model showed an area of 581 km² (35%) within the study area has high to good bear-occurrence probability values; 86% of this area is located in 2 patches, each larger than the average bear home range. Identification of these patches may support establishment of a reserve in the area, which would ensure long-term survival of the brown bear and sustainable water use and resource extraction from Pistacia atlantica forests by resident and nomadic communities in the region.