A study was conducted to identify habitat characteristics associated with age 0+ White Sturgeon (Acipenser transmontanus Richardson, 1863) recruitment in three reaches of the Columbia River Basin: Skamania reach (consistent recruitment), John Day reach (intermittent/inconsistent recruitment), and Kootenai reach (no recruitment). Our modeling approach involved numerous steps. First, we collected information about substrate, embeddedness, and hydrodynamics in each reach. Second, we developed a set of spatially explicit predictor variables. Third, we built two habitat (probability) models with Skamania reach training data where White Sturgeon recruitment was consistent. Fourth, we created spawning maps of each reach by populating the habitat models with in-reach physical metrics (substrate, embeddedness, and hydrodynamics). Fifth, we examined model accuracy by overlaying spawning locations in Skamania and Kootenai reaches with habitat predictions obtained from probability models. Sixth, we simulated how predicted habitat changed in each reach after manipulating physical conditions to more closely match Skamania reach. Model verification confirmed White Sturgeon generally spawned in locations with higher model probabilities in Skamania and Kootenai reaches, indicating the utility of extrapolating the models. Model simulations revealed significant gains in White Sturgeon habitat in all reaches when spring flow increased, gravel/cobble composition increased, or embeddedness decreased. The habitat models appear well suited to assist managers when identifying reach-specific factors limiting White Sturgeon recruitment in the Columbia River Basin or throughout its range.