A state-of-the-art optimization model was developed for prioritizing investments in culvert mitigation actions within the Big Lake area of Alaska. Unlike existing prioritization models, the model takes into account the spatial distribution of key habitats required throughout the full coho salmon life-cycle, the dispersal capabilities of fish, and the upstream/downstream passability of barriers. The model represents a radical improvement over the variety of ad-hoc methods commonly used in barrier prioritization planning (i.e., scoring and ranking procedure) and even existing optimization approaches aimed at improving connectivity for migratory fish. At present, just under half of river habitat in the Big Lake basin is currently available for meeting the life-cycle needs of coho. Access to winter rearing habitat (from age 0 summer rearing areas) limits connectivity the most (30% reduction in connectivity), followed by spawning grounds (17% reduction) and age 1+ summer rearing (15% reduction). Age 0 summer rearing and outward smolt migration do not have any substantial impact on connectivity. To increase available habitat to 100% would require the removal of 29 out of 60 existing culverts in the Big Lake basin at an estimated cost of approximately $6.8M. A 50% in available habitat can be achieved with a budget of just over$3M. Certain high-frequency culverts (those selected a high proportion of time by the optimization model) have lower than average passability for juveniles and adults, very large amounts upstream spawning, summer (for age 0 and 1+ juveniles), and winter rearing habitat, and significantly higher than average mitigation costs compared to culverts as a whole. These are mostly located near key strategic lakes or on mainstem stretch of river, which form major thoroughfares for coho dispersal.