Abstract: An optimal operation model for hydropower station is characterized by high-dimension, nonlinearity, multiple constraints, and difficult model solution. To surmount these problems, this paper presents a social spider optimization (SSO) algorithm to solve such models. This new algorithm searches the global optimum using the synergy mechanism of social spiders, so that it can avoid premature convergence and local-optimum trapping and obtain stable results for optimal decision on reservoir operation. Its performances in a case study are compared with those of the dynamic programming (DP) and genetic algorithm (GA). The comparison shows that with a smaller number of parameters, it is superior in calculations and more robust and efficient in optimum searching. Therefore, the social spider optimization algorithm is an effective method for practical optimization of hydropower station scheduling models.