River regulation, especially from dam operations, has serious impacts on downstream aquatic ecosystems, especially fish habitats. It is important to quantitatively evaluate these effects and to explore possible remedial measures. Ecohydraulic models that combine biotic (ecological) and abiotic (hydrodynamic) processes, have proved to be effective approaches. This study developed a Fuzzy Logic (FL) fish habitat model based on a series of field surveys and lab experiments. In order to optimize the membership functions and define appropriate fuzzy rules, a Genetic Algorithm (GA) was adopted. The model developed was applied to a channel of the Yalong River in the Southwest China, which will be affected by two cascading hydropower projects. Compared to the natural condition, the model results indicated that the reservoir operation clearly influenced the habitat of the target fish. A relationship between river discharge and habitat suitability index was investigated. In order to balance hydropower generation and ecological flow demand, it was proposed to ensure flow releases to maintain 70% of the total original habitat. The research also demonstrated that the application of GA to the FL model can certainly improve the model performance.