In the last few decades, discontinuum (or discrete, discontinuous) numerical modelling strategies-i.e. those capable of representing the motion of multiple, intersecting discontinuities explicitly-have become increasingly popular for the structural and seismic assessment of unreinforced masonry (URM) structures. The automatic recognition of new contact points and prediction of large deformations up to complete separation are unique features of discontinuum-based models, making them particularly suitable for unit-by-unit simulations. The adaptation of discrete computational models, primarily used for analyzing rock mechanics and geomechanics problems, to the conservation, structural and earthquake engineering evaluation of URM assemblies is still ongoing, and recent advances in computer-aided technologies are accelerating significantly their adoption. Researchers have now developed fracture energy-based contact models tailored to unreinforced masonry mechanics , explored discontinuum analysis from the mortar joint-to the 3D building-level, combined discrete modelling strategies with analytical or continuum approaches, integrated the latest structural health monitoring and image-based developments into discontinuum-based analysis framework. Concurrently, new and still unsolved issues have also arisen, including the selection of appropriate damping schemes, degree of idealization and discretization strategies, identification of appropriate lab or onsite tests to infer meaningful equivalent mechanical input parameters. This paper offers to the research and industry communities an updated critical appraisal and practical guidelines on the use of discontinuum-based structural and seismic assessment strategies for URM structures, providing opportunities to uncover future key research paths. First, masonry mechanics and discontinuum-based idealization options are discussed by considering micro-, meso-and macro-scale modelling strategies. Pragmatic suggestions are provided to select appropriate input parameters essential to model masonry composite and its constituents at different scales. Then, discontinuum approaches are classified based on their formulation, focusing on the Distinct Element Method (DEM), Applied Element Method (AEM) and Non-Smooth Contact Dynamics (NSCD), and an overview of primary differences, capabilities, pros and cons are thoroughly discussed. Finally, previous discontinuum-based analyses of URM small-scale specimens, isolated planar or curved components, assemblies or complex structures are critically reviewed and compared in terms of adopted strategies and relevant outcomes. This paper presents to new and experienced analysts an in-depth summary of what modern discontinuum-based tools can provide to the structural and earthquake engineering fields, practical guidelines on implementing robust and meaningful modelling strategies at various scales, and potential future research directions.