In setting-up a breakwater test in a laboratory flume, the conventional practice is to scale geometrically the armour units, underlayer rocks and core material based on the Proude scaling criteria. However, because some of the properties of the water are not scaled for a model test, the Reynolds scaling law is violated. This can result in improper water flow distribution through the model breakwater. To investigate this problem, a series of tests were performed with a model breakwater in which the permeability (and porosity) of the core was varied over a wide range, and the hydraulic response and breakage of the armour units were measured. It was found that the overall stability of the breakwater could be drastically affected if the flow in the core is scaled incorrectly.