Natural resource management over broad, heterogeneous landscapes is complicated by the inherent uncertainty in ecosystem processes and the ability of managers to predict the response of systems to management actions. Management decision-making is further complicated when multiple user-groups and management agencies with overlapping jurisdictions have fundamentally different objectives and policies. In these instances, formal decision making frameworks, such as structured decision making (SDM), can provide a means to evaluate management decisions in an integrated framework that can be used to address conflicting perceptions of system dynamics. In Alaska, brown bears (Ursus arctos horribilis) occur in large numbers on lands managed by the National Park Service (NPS) and other federal agencies and also are managed by the Alaska Department of Fish and Game and regulated as a game species by the Alaska Board of Game. Meanwhile, sea otter monitoring efforts in southwest Alaska are largely implemented by the National Park Service while the US Fish and Wildlife Service is the agency tasked with making decisions regarding sea otter management. Using SDM, we developed integrated modeling and decision support systems to explicitly link management, research, and monitoring of brown bears and sea otters in Alaska. The brown bear decision models tracked the state of bears through time in Katmai National Park and Preserve and Noatak National Preserve and estimated the effects of management actions on bear populations, harvest success, human-bear incidents, and park visitation. Sensitivity analysis identified key uncertainties that included factors that affected bear populations and human-bear incidents. In addition to eliciting values from decision-makers, benefit transfer was used as an alternate means of estimating values associated with fundamental objectives. This approach suggested that decision-makers values reflected the publics non-consumptive use and harvest values but that the value they placed on the bear population objective may have been too high. The model estimates also were sensitive to the relative value of harvest, bear population, and non-consumptive use objectives. Limiting the scope of the problem to NPS jurisdictional boundaries allowed for transparent decision making but may slow learning in an adaptive management framework.