The problem of the influence of power fluctuations of wind farms due to the variability of the wind speed on the electric power system is considered. With high wind energy penetration, an increase in the operating reserve in electric power systems is required to cover possible sudden power fluctuations. One of the ways to reduce the stochastic nature of the wind farms power generation is their geographically distributed location. A method is proposed for the selection of capacity and distributed placement of wind farms, taking into account the factor of the variability of the total generated power. In each of the prospective areas for wind farm placement, the simulation of electricity generation by a single wind turbine with hour-by-hour breakdown is carried out using the developed WindMCA software based on long-term ground-based weather stations data. Optimization of wind farms capacity and their distributed placement in areas is carried out using a genetic algorithm in the MATLAB environment. The target function is the coefficient of variation of the power generated by all wind farms in the areas under consideration, depending on the number of wind turbines therein. Power duration curves are used in the final comparison of wind farms siting options. The application of the method is carried out on the example of the wind farms placement in the Zabaykalsky Krai. A solution has been obtained that provides a minimum coefficient of variation of the wind farms generated power and a relatively high capacity utilization factor. With a distributed location of wind farms, the duration of the period with the maximum output is reduced, however, the duration of low power generation is significantly increased. With an increase in the number of wind farms connected to various nodes of the electric power system, a certain guaranteed level of power generation can be obtained, which, ultimately, will reduce the required amount of the reserve of generating capacities.