We propose development of a new methodology that can be used by forest and fire managers and planners to consider the potential effects of all aspects of fire management (i.e., fuels reduction to post-fire restoration) on native stream fishes and their habitats. State-of-the-art fire-hazard prediction models will be coupled with erosion models to create watershed-scale sensitivity analyses of fuel-reduction effects on sediment supply to streams. Sediment inputs, including road-related erosion and drainage diversions, will be evaluated in terms of potential effects to aquatic habitats and water quality. Vulnerability and persistence of affected fish populations will be evaluated using computer-supported analytical approaches that consider habitat quality, size, and connectivity, tailored to the life-history attributes of the species concerned. Integration of these components will be achieved using Bayesian Belief Networks (BBN). This approach integrates existing knowledge and analysis methods to identify at-risk habitats, to assess the potential for management actions to affect those habitats, and to anticipate the consequences for T&E fish populations.