Significant wave height (SWH) in shallow waters is assessed by generating two wave hindcasts; the first uses ERA-Interim wind fields and the second one from ERA5 to quantify the improvement of the ERA5 surface winds on the SWH representativeness, both in deep and shallow waters along the Chilean coastline. Additionally, wind field predictions from the Global Forecast System (GFS) were used to assess the representativeness of shallow waters. Oceanographic buoys were used to validate SWH in deep waters, while Acoustic Doppler Current Profiler (ADCPs) was equipped to measure waves in shallow waters. Energy spectrums coupling Wavewatch III and Simulating Waves Nearshore (SWAN) models were transferred to evaluate the performance of shallow water simulations. In general, the SWH from both wave hindcasts showed good performance. Nonetheless, those forced by ERA5 presented a better qualitative comparison of sea state temporal variability, which increased the correlation coefficients (>0.9), coefficients of determination (>0.8), and minor errors (RMSE, MAE, and BIAS) compared to oceanographic buoys and ADCPs. Additionally, in simulations forced by GFS, the temporal variability of the waves in shallow waters was successfully reproduced. Nevertheless, an increase in the RMSE, MAE, and BIAS error was statistically verified compared to ERA-Interim and ERA 5.