The accurate modelling of overtopping of coastal defences by tsunami waves is of vital importance for the formulation of disaster management strategies. To improve knowledge of this phenomena the authors conducted experiments on coastal structure overtopping using bores that were generated by a dam-break mechanism. Three types of structures were tested, namely a coastal dyke, a wall, and a wall of infinite height. The results highlight the necessity to consider the energy present in a bore to determine if a structure will be overtopped or not. As a result of these experiments an empirical formula to determine the height of overtopping given the incident bore height and velocity was validated. The study highlights the importance of clearly modelling the velocity and Froude number of a tsunami. Such experiments should be conducted on rough beds, for which a suitable Manning's n seems to be around 0.06 sm^-1/3. The study also contrasted the results obtained to those of the ASCE7 method, and concludes that the Manning's n values recommended in ASCE7 are probably too low.