In the last two decades the performance of numerical wind-wave models has improved considerably. Several models have been routinely producing good quality wave estimates globally since the mid-1980s. The verifications of wind-wave models have mainly focused on the evaluation of the error of the significant wave height Hs estimates. However, for wave energy purposes, the main parameters to be assessed are the wave power Pw and the mean (energy) period Te . Since Pw is proportional to Hs 2 Tc , its expected error is much larger than for the single-wave parameters. This paper summarizes the intercomparison of two wind-wave models against buoy data in the North Atlantic and the Mediterranean Sea to select the most suitable one for the construction of an Atlas of the wave energy resource in European waters. A full verification in the two basins of the selected model—the WAM model implemented in the routine operation of the European Centre for Medium-Range Weather Forecasts—was then performed against buoy and satellite altimeter data. It was found that the WAM model accuracy is very good for offshore locations in the North Atlantic; but for the Mediterranean Sea the results are much less accurate, probably due to a lower quality of the input wind fields.