In the last decades increasing attention has been devoted to masonry structures both from researchers and professionals. This is due to the awareness of the great importance of these structures in the historical urban context, together with the great risk that they suffer in seismic areas. Growing success has been obtained by a simplified approach that models masonry walls through “equivalent” plane frames, with concepts and procedures drawn from the study of reinforced concrete and steel frames. In this approach, known as Equivalent Frame Method (EFM), each masonry resisting wall is modelled as a system of linear (frame) elements repre-sentative of the behaviour of finite portions of the wall (pier and spandrel panels). Up to now EFM has proven to be effective in the case of new buildings characterized by regu-lar geometrical configurations and with openings’ dimensions for which the frame-like as-sumption is suitable. Critical issues emerge from existing buildings in European historical centres, where irregularities are almost always present and geometrical anomalies can be detected even in the case of regular walls. For the specific case of geometrically regular URM walls, this paper presents sample cases tested with linear and non-linear analyses, with the aim to explore the applicability of EF procedures and to identify its limits. A comparison between an EFM procedure with a fiber approach and a more detailed FEM method with plane elements, is adopted as a validation tool, to evaluate the accuracy of the results provided by the EFM for walls characterized by different geometrical configurations.