Infill walls in frame structures greatly influence the dynamic behaviour of buildings and consequently their performance when under seismic loads. In building construction, unreinforced masonry walls constitute one of the oldest and worldwide used materials. Despite of great simplicity of their application due to their assemblage process, their behaviour under earthquake events is rather complex, and it is usually neglected in the design phase of building structures. Numerous studies have been conducted over the years to model and assess the influence of non-structural infill walls in buildings. Micro-, meso-and macro-models are available in the literature, although the choice of one of these types of models depends on the problem under consideration. For the current investigation, macro-models are more adequate, in particular equivalent strut models. This kind of modelling has been evolving over the years accounting for a large number of phenomena intrinsic to these elements: the failure modes expected by the infill walls; the relative infill-frame lateral stiffness; the consideration of openings in infills and partially infill walls; bounding frame columns' shear failure; vertical loading effect; out-of-plane behaviour of the infill and its interaction with in-plane behaviour; applicability to multiple degrees-of-freedom systems with multiple stories and bays, etc. This study considers two well-known macro-modelling strategies of unreinforced infill masonry walls in reinforced concrete frame structures modified to improve the in-plane and outof-plane behaviour simulation under seismic actions. Results showed that these modifications allowed for better reproduction of infill walls' behaviour compared with experimental tests.