This paper describes a linear programming model and Monte Carlo simulation which solve a middle-termed hydrothermal dispatch problem for three different scenarios of wind power penetration. Transmission line effects are neglected. The fuel cost in thermal plants varies in each time period. The test system has two thermal and three hydroelectric plants that can continuously operate during the twelve-months planning term. Wind generation is modeled by using Weibull Distribution Functions particular to the Colombian system, while hydro-generation is modeled by Normal Distribution Functions. Results show costs reduction for the three scenarios of wind power penetration. The main contribution of this paper is to simultaneously introduce the stochastic effect of wind and hydrology in the hydrothermal dispatch.