In this investigation, locations in the department of La Guajira, Colombia, with optimal wind speed behaviour for the generation of wind energy, were identified. The WRF meteorological model was used with input data from the GFS global analysis and forecast, two nested domains of 9 and 3 km, 45 vertical levels and the physical parameterization contained in the TROPICAL module. Successful model execution improved the spatial and temporal resolution of the input data. The resulting information corresponding to the wind speed was used to determine its behaviour at harvest heights of 50, 100 and 150 m. Subsequently, adjustment of the Weibull probability function was carried out, its shape and scale parameters were estimated via the maximum likelihood method, and finally, taking advantage of the cubic relationship between the wind potential and wind speed, the potential raster for the elevations was estimated. The values modelled by the WRF model were compared with the data recorded at the meteorological station located at the Bolívar airport at a height of 10 m above sea level, with an RMSE of 2.26 m/s, an R2 of 0.73 and a bias of 1.17 m/s. The shape and scale parameters in the modelled data are overestimated by 0.79 and 1.83 m/s, respectively, compared with the fit of the values recorded by the weather station. In terms of potential, the greatest resource is found over the sea, mainly in areas near the Sierra Nevada and in the northern part of the department, reaching values of up to 1200 W/m2.