The paper presents a method to analyse wind data for wind-turbine siting. Two parameters are defined by means of which the siting can be optimised, the site effectiveness (ratio output/available energy) and the cut-in speed. It is shown that the site effectiveness achieves a maximum for a value of the cut-in speed which depends on the rated speed and on the site. The cut-in speed of a WECS should be chosen according to the rated speed of the turbine and the cubic mean wind speed of the site. The wind data of two regions (Somalia and Calabria, Italy) with very different climates are examined and compared, showing correlation with the cubic mean wind speed. It is also shown that the Weibull and Rayleigh wind-frequency distribution models are not reliable for evaluating the output energy of a WECS.