Probability density functions, such as the Weibull and Rayleigh, are often used in wind speed and wind energy analyses. It is important to establish the goodness of their suitability in modelling a particular measured distribution. The parameters obtained from the distributions (distributional parameters) are required to assess the suitability of the functions. Such distributional parameters that are commonly used are the following: the parameters of the function itself (such as the parameters of the Weibull function), the wind energy output and the correlation coefficient of the probability density distributions. It is shown in the present article that these parameters alone are not sufficient to determine the suitability of a given function. Additional parameters are required to have a complete statistical assessment of the function, such as the standard deviation or, alternatively, the correlation coefficient of the power density distributions. In the present study, the suitability of the Weibull and Rayleigh functions is assessed based on a total of six parameters (of the probability density and power density distributions), calculated from 12 months of hourly time series wind speed data. It is shown that the degree of suitability of a given function changes to a great extent with the parameter on which the judgement is based. (C) 2003 Elsevier Ltd. All rights reserved.