The wind speed distribution is the basis for the assessment of wind energy potential needed for the design of wind farms. Therefore, the proper specification of the wind speed distribution is of special importance. In this study, we propose for the first time two flexible families of distributions, the skewed generalized error and skewed t, for the description of the wind speed distribution. These families both are very flexible enough to accommodate the shape of the wind speed data and include some well-known distributions as special cases. Also we evaluate the performance of these flexible families relative to the widely-used Weibull distribution by using wind speed data measured in various regions of Turkey. The results indicate that these flexible families of distributions provide substantial improvement over the Weibull distribution in estimating wind speed distribution and wind power density distribution. Thus, the skewed generalized error and skewed t distributions can be alternatively used for assessment of wind energy potential. (C) 2011 Elsevier Ltd. All rights reserved.