The Skagit Chinook Recovery Plan provided estimates of new annual parr migrant production from proposed floodplain restoration projects in the Skagit River, Washington, an important watershed for threatened Chinook Salmon. Until now, tools have been limited to link restoration planning and design with juvenile Chinook Salmon abundance in ways that allow comparison among project alternatives, across sites, and between observed fish use and the Skagit Chinook Recovery Plan. Occupancy models have shown utility in restoration planning in other basins like the Trinity River, California. We performed cross-validation and occupancy threshold selection on grid-based fish habitat occupancy models developed in Barnaby Slough to calculate juvenile Chinook Salmon abundance. We then applied the results in ten scenarios of season, site, and restoration time period to validate the models and assess the applicability of the tool in areas outside the original model domain. We found that parr migrant estimates were substantially higher than combined-season parr/yearling estimates, corresponding with known biology. For combined-season parr/yearling, fish habitat model calculated abundances were lower than those in the Recovery Plan, likely due mostly to low incidence of yearling life histories in the watershed. However, several other factors could contribute, including limited anticipated changes in site depth and velocity that would generate new fish habitat and a reduction in the baseline Chinook Salmon outmigrant abundance since the time of Recovery Plan writing. Additional work is needed to explore the relative role of these factors and to solidify links between grid-based fish habitat models and Recovery Plan estimates at the annualized scale but this tool provides new ability to discriminate among restoration projects and alternatives in terms of Chinook Salmon benefit.