As the installed capacity of wind power in the SEIN (National Interconnected Electric System) of Peru increases (developing country), very useful predictors for developing certain wind resource are obtained to provide stability to the system. The performance of WRF (Weather Research and Forecasting - version 3.4) forecast winds was evaluated using different parameterization options in the department of Ica (Peru) to minimize the uncertainty of the prediction. The complexity of the terrain and the resolution of the selected domain aspects were considered in the model. The wind rose and the frequency histograms were used to evaluate the model and it was noted that the wind frequency for low and intermediate speeds was underestimated. The wind direction was simulated in a similar manner but specifically for dominant wind regimes. For low wind speeds, the characterization of wind direction is very subjective with large deviations obtained between the simulation alternatives. The best numerical configuration is one that does not exceed 24 h: but where the error increases with the complexity of the terrain leading to the lack of accurate data. The wind direction is simulated by the model especially with strong dominant sectors. In January, it was detected significant frequencies in low speeds (1-2 m/s). In these cases, the characterization was subjective and could be higher the deviation between observed and simulated results. In June, the simulated wind speeds are constant below 4 m/s and above 11 m/s, obtaining similar results in dominant wind sectors. Increasing the resolution of the simulation domain was not sufficient to improve the efficiency of the model. (C) 2014 Elsevier Ltd. All rights reserved.