Among all uncertainty factors affecting the wind power assessment at a site, wind speed extrapolation is probably one of most critical ones, particularly if considering the increasing size of modern multi-MW wind turbines, and therefore of their hub height. This work is intended as a contribution towards a possible harmonisation of methods and techniques, necessarily including surface roughness and atmospheric stability, aimed at extrapolating wind speed for wind energy purposes. Through the years, different methods have been used to this end, such as power law (PL), logarithmic law (LogL), and log-linear law (LogLL). Furthermore, aside from applying PL by using a mean wind shear coefficient observed between two heights ( α ¯ ), a number of methods have been developed to estimate PL exponent α when only surface data are available, such as those by Spera and Richards (SR), Smedman-Högström and Högström (SH) and Panofsky and Dutton (PD). The main purpose of this work is to analyse and compare the skill of some of most commonly used extrapolation methods once applied to a case study over a coastal location in Southern Italy. These are LogLL, LogL, as well as PL by using different approaches to estimate α (i.e., PL- α ¯ , PL–SR, PL–SH, and PL–PD). In doing so, the influence of atmospheric stability and surface roughness ( z 0 ), with special attention to their variability with time and wind characteristics, has been also investigated. In addition, a comparison among the three α-estimating methods by SR, SH and PD has been carried out. A 6-year (1997–2002) 1-h meteorological dataset, including wind measurements at 10 and 50 m, has been used. In particular, the first 5 years were used to analyse site meteorology, stability conditions, and wind pattern, derive α and z 0, as well as compare α-estimating methods, while the latter (2002) to test the skill of the extrapolation methods. Starting from 10-m wind speed observations, the computation of 50-m wind speed and power density, as well as wind resource and energy yield, has been made. The Weibull distribution and related parameters have been used for the wind resource assessment, while AF, CF and AEY were calculated to evaluate the potential wind energy yield.