The influence of meteorological variables on measured power curves of multi megawatt wind turbines (WT’s) has been investigated in the frame of certification measurements. The three investigated WT’s are of type Enercon E112 (6.0 MW and 4.5 MW rated power, 124 m hub height) and Enercon E-70 E4 (2.3 MW rated power, 65 m hub height) and are located in flat terrain. At all three machines large effects of the turbulence intensity are observed. The measured power coefficients increase significantly with increasing turbulence intensity below rated power and decrease significantly with increasing turbulence intensity in the transition region to rated power. The effect has the same magnitude at all three investigated WT’s. After isolating turbulence and wind shear effects, nearly no effect of the vertical wind shear remains at the Enercon E112 machines. At the Enercon E-70 E4 a significant wind shear effect is observed with the power coefficient decreasing with increasing wind shear below rated power. This is believed to be due to the low hub height of the investigated turbine. A new approach for normalising the power curve data to a target turbulence intensity in respect to the effect of averaging the power curve data over 10-minutes has been developed. This Normal Distribution Model is capable to model the time averaging effect also at high turbulence intensities and in the transition region to rated power. It turns out that only a small part of the observed turbulence effects can be explained by the time averaging effect.