Within an existing transmission network, this paper considers the problem of identifying the wind power plants to be built by a wind power investor to maximize its profit. For this analysis a future target year is considered and the loads at different buses are represented by stepwise load-duration curves. The stochastic nature of both load and wind is represented via scenarios. The considered electric energy system operates under a pool-market arrangement and each producer/consumer is paid/pays the Local Marginal Price (LMP) of the bus at which it is located. The higher the wind penetration is, the lower the resulting LMPs. To tackle this problem a stochastic bilevel model is proposed, whose upper-level represents the wind investment and operation decisions with the target of maximizing profits; and its lower-level represents the market clearing under differing load and wind conditions and provides LMPs. This model can be recast as a mixed-integer linear programming problem solvable using commercially available branch-and-cut solvers. The proposed model is illustrated using an example and two case studies. (C) 2011 Elsevier Ltd. All rights reserved.