The behaviour of slender concrete masonry walls is poorly understood. This is particularly evident by the treatment of such walls in the Canadian masonry design standard, CSA S304-14. Many researchers have noted the poor accuracy of CSA S304-14’s provisions and the need for more experimental data. A major limiting factor to building and testing walls is their size and the quantity of materials used. This limitation can be overcome by using half-scale walls – however, it must first be proven that scale effects do not influence the results obtained. Of further utility would be a verified finite-element model to predict wall failure loads and modes with a high degree of confidence. Towards these ends, twelve masonry walls, six at full-scale and six corresponding half-scale walls, were built and tested under an axial load with an eccentricity of 1/3 of the wall thickness;​ single curvature moment was applied. Digital image correlation (DIC) was used to acquire lateral displacement data;​ its implementation in this study is presented. Specimen handling procedures are also presented. Wall strengths, lateral displacement profiles and behaviour were recorded and are presented. Experimental results show that size effect does not influence the behaviour and strength of slender concrete masonry walls at half-scale. Experimental results are used to discuss shortcomings in CSA S304-14’s treatment of wall effective stiffness. A verified finite element model was also used to investigate the failure load and mode of 80 hollow masonry walls, with varying effective height to thickness ratio, eccentricity, and inclusion of initial geometric imperfection. Numerical results are used to determine when the inclusion of geometric imperfections impacts numerical predictions, and to determine wall failure modes for various parametric combinations. The model is also used to discuss adjustments to clauses in CSA S304- 14 related to minimum primary moment and hollow masonry wall axial strength.