Recently, cogeneration units have played an increasingly important role in the utility industry. Therefore the optimal utilization of multiple combined heat and power (CHP) systems is an important optimization task in power system operation. Unlike power economic dispatch, which has a single equality constraint, two equality constraints must be met in combined heat and power economic dispatch (CHPED) problem. Moreover, in the cogeneration units, the power capacity limits are functions of the unit heat productions and the heat capacity limits are functions of the unit power generations. Thus, CHPED is a complicated optimization problem. In this paper, an algorithm based on Benders decomposition (BD) is proposed to solve the economic dispatch (ED) problem for cogeneration systems. In the proposed method, combined heat and power economic dispatch problem is decomposed into a master problem and subproblem. The subproblem generates the Benders cuts and master problem uses them as a new inequality constraint which is added to the previous constraints. The iterative process will continue until upper and lower bounds of the objective function optimal values are close enough and a converged optimal solution is found. Benders decomposition based approach is able to provide a good framework to consider the non-convex feasible operation regions of cogeneration units efficiently. In this paper, a four-unit system with two cogeneration units and a five-unit system with three cogeneration units are analyzed to exhibit the effectiveness of the proposed approach. In all cases, the solutions obtained using proposed algorithm based on Benders decomposition are better than those obtained by other methods. (C) 2013 Elsevier Ltd. All rights reserved.