A methodology is developed to optimize the size and the location of power plants supplying given demand centers by minimizing the cost of transmission lines and plant capital costs subject to the physical constraint that the power plants must be located within a predetermined feasible geographical region. The optimization problem falls within a class of mixed integer nonlinear constrained programming for which no general method of solution exists. Optimization is carried out in two steps to separate considerations of integer and continuous variables. A complete set of possible configuration alternatives in terms of the number of power plants is first generated by examining the comers of a polyhedron set defined by the upper and lower bounds on the number of power plants at each location, with the demand satisfied through a predefined directed transmission network. Then, through a constrained nonlinear programming technique, the optimum location for each promising, feasible alternative is calculated. The best alternative, i.e., the one having the minimum total cost, is selected as the optimum solution.