Wind turbines installed in a wind farm can be affected by wake from neighboring wind turbines, which reduces power production and shortens turbine life due to mechanical fatigue. Therefore, careful consideration of the wake effects is necessary before establishing an optimum layout design for wind farms. Until now, various numerical models have been developed and applied to evaluate the wake effects, but comparison among (and verification of) the measurement results conducted at numerous wind farms have been few and insufficient. To verify the prediction accuracy of engineering wake models, namely, eddy viscosity, Larsen, Jensen, and Frandsen models, which are widely used in wind energy business, the current study presents the results of the comparative analysis of the values measured at a commercially operated onshore wind farm. The results demonstrate that the Jensen model is the best model in predicting the wake-centered velocity deficit under a specific wind-speed condition, and the eddy viscosity and Larsen models are relatively accurate in predicting the width of the wake and its profile. In conclusion, we need to carefully apply a model for wake-effect assessment based on the spacing between wind turbines because the prediction accuracy of the wake models varies with the downstream distance condition. (C) 2015 Elsevier Ltd. All rights reserved.