Detailed testing of a new two‐dimensional hydraulic modelling system is presented. The methodology consists of applying the hydraulic model to a set of theoretical tests, for which analytical solutions are known, and then comparing a model simulation with a real flood event. The water‐at‐rest test, three dam‐break tests and the Seiche test constitute the set of tests with analytic solutions. The flood event is the Boscastle 2004 flood, for which observed water levels are available. The model yields results that compare closely with the analytical solutions in all of the tests, except for the Seiche test. When applied to the real flood event, the simulated maximum levels are close to the observed levels. Comparisons with two similar commercial software packages show similar results between the models.