The confined masonry typology has been traditionally used for building, and also selected for the reconstruction of recently earthquake-damaged cities, in developing countries responding to the seismic-economic couple. However, most of the procedures in design codes adopted for these countries are force-based, which appear to be inadequate for loading cases under severe earthquakes, where the response in displacement plays the essential role for ultimate (life preservation) limit state. In this work, a worldwide review is made of the experimental response of confined masonry buildings, from which a first storey, shear-dominated, mechanism is mostly identified. Then, two full-scale confined masonry structures with regular and irregular plan configurations, quasi-static tested, are analysed under push-over loading of simplified models of the buildings. The idealized models are based on the use of frame and discrete spring elements, allowing to consider the interaction between the masonry panel and the r.c. confining elements. A comparison between the results of tests and the analytical predictions is made, particularly concerning the base shear-displacement response and the damage patterns. The accuracy of the predictions is very satisfactory, allowing to capture the base shear-displacement response envelope and the general damage trend on the buildings, and thus making the method able for performance-based design procedures.