The CLASH project (Crest Level Assessment of Coastal Structures by Full Scale Monitoring, Neural Network Prediction and Hazard Analysis on Permissible Wave Overtopping, www.clash-eu.org) funded by European Union (EU) through contract no. EVK3-CT-2001- 00058, investigated wave overtopping for different structures in prototype and in laboratory (see De Rouck et al., 2002). The main scientific objectives of CLASH were (i) to solve the problem of possible scale effects of wave overtopping and (ii) to produce guidelines for crest height design or assessment, based on overtopping criteria. This report is a synthesis of the project results. The CLASH methodology to estimate mean overtopping discharges and overtopping risk assessment is based on four elements: (1) the CLASH database (workpackage 2: WP2), (2) the hazard analysis including socio-economic impact (WP6), (3) the scale and model effect analysis (WP7) and (4) the NN generic prediction method (WP8). For any given specific engineering application, hazard analysis provides the limit of permissible overtopping discharges for different modes of failure or damage (pedestrians, vehicles, property, etc.). Scale and model effect analysis defines the relationship between prototype overtopping and Froude scaled overtopping. Finally, the NN generic prediction method, based on the CLASH database, estimates the overtopping discharges corresponding to the Froude scaled model. The CLASH executable file NN-OVERTOPPING 2.0 is designed for end-users to calculate NN Froude scaled mean overtopping discharges and confidence intervals. In addition prototype mean overtopping estimations, considering scale and model effects are provided. These overtopping estimations can be compared to permissible overtopping for the damage risk assessment (overtopping hazard analysis).