Design of breakwaters is largely based upon guidelines developed from model tests and prototype experience of non-overtopped structures. It is generally recognised that an overtopped breakwater design according to these procedures may be unsafe and that special attention needs to be paid to the crest and leeward face. In practice this is achieved by rigorous model testing of each design. We attempt to provide more realistic design guidelines from the primary variables. The study was only partially successful., and more questions were raised than answered. The parameters influencing the stability of an overtopped structure are far more complex than of a non-overtopped. Design guidelines based on the latter have little relevance, and often may lead to unsafe design.