In the study, the performance of the MIKE 21 SW (Spectral Wave) model using the three different wind fields, namely ERA-Interim, ERA5, and CFSR was evaluated in the Eastern Mediterranean Sea. Model results were calibrated with four buoy measurements by tuning physical model parameters. Wave simulations showed a strong sensitivity to the whitecapping parameter (Cds). Calibrated MIKE 21 SW model run to validate at two buoy measurements. Considering the statistical error measures: (i) ERA-I predicted lower significant wave heights and wave periods than ERA5 and CFSR, (ii) statistical error measures (bias, RMSE, and SI) obtained with CFSR were calculated slightly higher than those obtained using ERA5 (iii) ERA5 performed slightly better in the hindcast of Eastern Mediterranean wave properties than ERA-I. As a result of all evaluations, the highest correlation coefficient (R), relatively low statistical error measures, and slightly better accuracy were provided by the model settings forced with ERA5 wind fields.