The shove test (ASTM Standard C1531) is an experimental technique aimed at studying the shear-sliding behavior of brick masonry. It can be executed according to various testing methods that differ in the way the vertical compression load is applied and in the way bricks and/or joints are locally removed for inserting jacks. One of the most critical aspects is the correct evaluation of the compressive stress state on the sliding brick. The objective of the present paper is to investigate the capability of the shove test in determining the shear strength parameters of brick masonries and to highlight the main advantages and disadvantages of the various testing methods. To this aim, nonlinear numerical simulations of the shove test were performed by adopting a brick-to-brick modeling strategy. The 2D numerical model was calibrated and validated through comparisons with experimental results of triplet tests and shove tests. The numerical analyses allowed to understand the influence the different testing methods and the masonry mechanical properties, such as dilatancy, may have on the test results. Based on the numerical outcomes, correction factors were calibrated for the proper evaluation of the compressive stress state on the sliding brick. Improvements with regards to the experimental procedures, i.e. additional test phases and measurements, were also proposed to enhance the results interpretation.