The fish in the dam reservoir will die from the strike of rotating runner blade with high speed when passing the turbine flow passage. In order to improve the survival rate of the fish, a method for evaluating the strike mortality of fish in Francis turbine runner was presented. This method contained three steps. Firstly, the CFD method was adopted to obtain the velocity field inside the Francis turbine runner. Secondly, based on the blade-strike model and the absolute velocity distribution on the inlet of runner, the probability of fish struck the runner blade was calculated. Finally, according to the fitting model of radial basis function (RBF) neural network, the lethality rate of fish owing to blade strike was obtained, and then the strike mortality of fish passed through the runner could be calculated. Based on this method, the strike mortality of fish of the initial and the optimized runner at different operation points were calculated. The influence of the turbine discharge and blade numbers on the mortality rate of fish was researched. From the result, it can be concluded that the mortality rate of fish passed through the runner was positively correlated with the discharge and the number of runner blades under the rated head. For the initial runner, the strike mortality rate of fish reached 16.85% at rated point. Compared with the initial runner, the strike mortality rate of fish at rated point was reduced by 2.93 and 5.3 percentage points, which meant these two optimized runner had better ecological performance of fish than the initial runner. According to the comparison of ecological performance and hydraulic efficiency, the optimized runner with 13 blades was chosen as the final scheme. The research methods and contents can provide valuable reference for the fish friendly optimization of Francis turbine.