In this chapter, a new methodology to manage coastal protection by means of wave farms is proposed. Artificial intelligence tools, more specifically artificial neural networks (ANNs), were used to assess dry beach surface differences between the no wave farm situation and different wave farm project scenarios. A number of alongshore locations and layouts—represented as the number of rows and the spacing between devices—formed the wave farm scenarios and the influence of the wave climate—significant wave height and mean wave direction—was also taken into account. The selected study site was Playa Granada (southern Iberian Peninsula), a beach with important erosion problems. The datasets used for training and testing the ANN were obtained by means of a suite of numerical models including a third generation wave propagation model, a sediment transport formulation and a shoreline response equation. In order to obtain the ANN which provides the best fit to the data, a comparative study involving more than forty different architectures formed by one and two hidden layers and trained by means of two training algorithm was carried out. The [5-10-1] architecture obtained the best results with a correlation coefficient and RMSE of 0.9489 and 4.22 m22, respectively. Once the best architecture was found, the ANN was applied to the study site in order to obtain the optimum location and layout for a wave farm project and the results indicate that dry beach surface could increase up to 5400.18 m22 per year. These results show that ANNs can be useful for managers to optimize the design of wave farms for coastal protection.