Significant effort has been made to generate a homogeneous database on wave overtopping consisting of more than 10,000 irregular wave overtopping tests from more than 160 independent projects or test series, each described by means of 31 parameters. Many coastal structures, including dikes, rubble mound breakwaters, berm breakwaters, caisson structures and combinations have been considered and have been schematised for inclusion in the database. All these overtopping tests are represented by over 300,000 numbers in the database. Within the framework of the CLASH-project specific overtopping tests on missing parameters and structures (white spots) in the database have been performed, as well as prototype overtopping measurements and small scale overtopping simulations. The results of these tests have enlarged and homogenized the database. White spot tests concentrated on the roughness of all kind of armour units and on the influence of oblique wave attack. These tests were performed by the University of Edinburgh (UK) and Aalborg University (DK), respectively. Prototype measurements were performed at Ostia (IT), Samphire Hoe (UK) and Zeebrugge (BE). The small scale simulation tests were performed by most of the partners in CLASH. The database has been created in such a way that particular and confidential information on actual projects is not present. This creates the opportunity to finally release the database to all interested people. The database was created as a first step and as basis for the generic prediction method on wave overtopping. The latter has been described in Van Gent [Van Gent, M.R.A., Van den Boogaard, H.F.P., Pozueta, B., Medina, J.R., 2007. Neural network modelling of wave overtopping at coastal structures, Coastal Engineering, Vol. 54, Issue 8, pp. 586–593]. But the database can also be applied on its own. Furthermore, the minimum set of hydraulic and structural parameters to describe a test is very useful in creating other databases and could be considered as essential parameters to be described in each test report on physical model investigation of coastal structures. The paper describes: • the gathering of the data • the screening of the data • the development of parameters to describe each individual test • the parameters required for an artificial neural network prediction method • a parameter analysis of the total database • the application of the database for design purposes