The tolerances as applied by the rotor blade and turbine manufacturers for a finished blade were investigated, discussed and harmonized. Due to the influence of the rotor design, e.g. tip speed ratio, power control and blade sections used, general tolerances applicable for all blades are limited. The participants agreed on a minimum set of tolerances. Blade section and mould coordinates at app. 0.71, 0.84 and 0.95 rotor radius were measured by DEWI at six blades and four types of serial produced 20m blades. The aerodynamic properties were simulated and compared to the reference blade section. For two blades the rotor power curve and energy output were calculated and compared to the reference blade. Possible ageing effects were investigated by measuring a one year old blade. The results show that today's aerodynamic blade quality is fair or even good. Nevertheless improvements are possible, especially in the leading edge region. The typical energy loss due to blade section tolerances seems to be 1-2%, based on the limited number of blades investigated. Recommendations for the rotor blade quality control have been derived from the discussion on rotor blade tolerances and the results of the blade sections measured. Four 40m prototype blades of the NÄSUDDEN II and AEOLUS II wind turbines, that were built in the same mould, were measured by FFA. Blade sections and twist angles were directly measured at the turbine, using special equipment. Significant differences were found, compared to the intended airfoils, as well as between the four blades. Nevertheless the energy output is reduced by 1.5% only, the same result as above. The influence of the blade manufacturing errors, including twist deviation and blade section shape errors, on the aerodynamic performance of a horizontal axis wind turbine has been simulated by CRES. The adopted numerical method is based on stochastic analysis and standard blade element theory. It is shown that the standard deviations of the rotor power and thrust coefficients are proportional to the standard deviations of the introduced error on the twist angle and the aerodynamic parameters employed. The latter are defined by parametrizing the characteristic lift and drag curves of the blade sections. The global effect on power production and blade loading can be obtained through the linear superposition of the individual effects, as long as the introduced errors are small.