The impact of atmospheric stability on vertical wind profiles is reviewed and the implications for power performance testing and site evaluation are investigated. Velocity, temperature, and turbulence intensity profiles are generated using the model presented by Sumner and Masson. This technique couples Monin-Obukhov similarity theory with an algebraic turbulence equation derived from the $k-ϵ$ turbulence model to resolve atmospheric parameters $u*$, $L$, $T*$, and $z0$. The resulting system of nonlinear equations is solved with a Newton-Raphson algorithm. The disk-averaged wind speed $u¯disk$ is then evaluated by numerically integrating the resulting velocity profile over the swept area of the rotor. Power performance and annual energy production (AEP) calculations for a Vestas Windane-34 turbine from a wind farm in Delabole, England, are carried out using both disk-averaged and hub height wind speeds. Although the power curves generated with each wind speed definition show only slight differences, there is an appreciable impact on the measured maximum turbine efficiency. Furthermore, when the Weibull parameters for the site are recalculated using $u¯disk$, the AEP prediction using the modified parameters falls by nearly 5% compared to current methods. The IEC assumption that the hub height wind speed can be considered representative tends to underestimate maximum turbine efficiency. When this assumption is further applied to energy predictions, it appears that the tendency is to overestimate the site potential.