The optimal operation scheduling of a pumping station with multiple pumps is formulated as a dynamic programming problem. Based on the characteristics of the problem, an extended reduced dynamic programming algorithm (RDPA) is proposed to solve the problem. Both the energy cost and the maintenance cost are considered in the performance function of the optimization problem. The extended RDPA can significantly reduce the computational time when it is compared to conventional DP algorithms. Simulation shows the feasibility of the reduction of the operation cost. (C) 2012 Elsevier Ltd. All rights reserved.