This paper presents new solution algorithms based on an evolutionary particle swarm optimization (EPSO) for solving the pumped-storage (P/S) scheduling problem. The proposed EPSO approach combines a basic particle swarm optimization (PSO) with binary encoding/decoding techniques as well as a mutation operation. The binary encoding/decoding techniques are adopted to model the discrete characteristics of a P/S plant. The mutation operation is applied to accelerate convergence and escape local optimums. The optimal generation schedules for both P/S and thermal units are concurrently obtained within the evolutionary. process of a scoring function. Therefore, hydrothermal iteration is no longer needed. The proposed approach is applied with great success to an actual utility system consisting of four P/S units and 34 thermal units. Experimental results indicate the attractive properties of the EPSO approach in a practical application, namely, a highly optimal solution and robust convergence behavior.