Based on the experience of recent violent earthquakes, the limits of the methods that are currently used for the definition of seismic hazard are becoming more and more evident to several seismic engineers. Considerable improvement is felt necessary not only for the seismic classification of the territory (for which the probabilistic seismic hazard assessment—PSHA—is generally adopted at present), but also for the evaluation of local amplification. With regard to the first item, among others, a better knowledge of fault extension and near-fault effects is judged essential. The aforesaid improvements are particularly important for the design of seismically isolated structures, which relies on displacement. Thus, such a design requires an accurate definition of the maximum value of displacement corresponding to the isolation period, and a reliable evaluation of the earthquake energy content at the low frequencies that are typical of the isolated structures, for the site and ground of interest. These evaluations shall include possible near-fault effects even in the vertical direction; for the construction of high-risk plants and components and retrofit of some cultural heritage, they shall be performed for earthquakes characterized by very long return periods. The design displacement shall not be underestimated, but neither be excessively overestimated, at least when using rubber bearings in the seismic isolation (SI) system. In fact, by decreasing transverse deformation of such SI systems below a certain value, their horizontal stiffness increases. Thus, should a structure (e.g. a civil defence centre, a masterpiece, etc.) protected in the aforesaid way be designed to withstand an unnecessarily too large earthquake, the behaviour of its SI system will be inadequate (i.e. it will be too stiff) during much more frequent events, which may really strike the structure during its life. Furthermore, since SI can be used only when the room available to the structure laterally is sufficient to create a structural gap compatible with the design displacement, overestimating this displacement may lead to unnecessarily renouncing of the use of such a very efficient method, especially in the case of retrofits of existing buildings. Finally, for long structures (e.g. several bridges or viaducts and even some buildings) an accurate evaluation of the possibly different ground displacements along the structure is required (this also applies to conventionally built structures). In order to overcome the limits of PSHA, this method shall be complemented by the development and application of deterministic models. In particular, the lack of displacement records requires the use of modelling, once they are calibrated against more commonly available velocity or acceleration records. The aforesaid remarks are now particularly important in the P.R. China and Italy, to ensure safe reconstruction after the Wenchuan earthquake of May 12, 2008 and the Abruzzo earthquake of April 6, 2009: in fact, wide use of SI and other anti-seismic systems has been planned in the areas struck by both events.