A mathematical model for vertical extrapolation of the measurement data for wind speed taken at several measurement heights is presented. The model is based on the method of least squares (LES). By applying the proposed model on the sets of measured data taken at least at three measurement heights, one obtains a synthetic set of data at a desired height where the wind power potential is analyzed. The basic idea is that during the process of estimation of the wind power potential the measurement data are first extrapolated by the proposed method and then by using program WAsP the spatial extrapolation is carried out. The algorithm is tested by one year wind speed measurement data taken at three locations characterized by different topographies of the terrain and different climatic conditions. The performed analyses show that pre-processing of measurement data by the proposed method results in a better estimate of the wind power potential at a height which is greater than the measurement heights compared to that obtained by the standard application of WAsP program which makes use of measurement data taken at one measurement height. (C) 2011 Elsevier Ltd. All rights reserved.