We used a spatially explicit population model of wolves (Canis lupus) to propose a framework for defining rangewide recovery priorities and finer-scale strategies for regional reintroductions. The model predicts that Yellowstone and central Idaho, where wolves have recently been successfully reintroduced, hold the most secure core areas for wolves in the western United States, implying that future reintroductions will face greater challenges. However, these currently occupied sites, along with dispersal or reintroduction to several unoccupied but suitable core areas, could facilitate recovery of wolves to 49% of the area in the western United States that holds sufficient prey to support wolves. That percentage of the range with recovery potential could drop to 23% over the next few decades owing to landscape change, or increase to 66% owing to habitat restoration efforts such as the removal of some roads on public lands. Comprehensive habitat and viability assessments such as those presented here, by more rigorously defining the Endangered Species Act's concept of “significant portion of range,” can clarify debate over goals for recovery of large carnivores that may conflict with human land uses.