Wind resource assessments often rely on models of wind-speed statistics using a Weibull distribution. This study shows how its use impacts the prediction of the wind energy content and the power output. Three methods for fitting the Weibull distribution are tested (maximum likelihood, moments and WAsP). The first two methods overestimate the production up to 5 %. The WAsP method always produces errors lower than 2 %. A Rayleigh–Rice distribution is also tested and shows even better skill.