<p>Three-dimensional (3D) magnetic inversion plays a critical role in magnetic exploration by providing information about the spatial location, geometric shape, and distribution of physical parameters of anomalous bodies. The size of the model space subdivision determines the inversion resolution, while there is currently no consensus regarding the selection of subdivision accuracy for model space, which hampers its practical application. By discussing the relationship between subdivision accuracy and the effectiveness of 3D inversion, this study aims to provide a basis for the selection of the size of the model space subdivision. Multiple sets of theoretical magnetic models are used for 3D inversion with different size of the model space subdivisions, and the correlation coefficient between the inverted magnetization model and the theoretical magnetization model is used to evaluate the inversion accuracy. The results showed that the inversion accuracy continuously increases as the model subdivision spacing decreases to 0.5 times of the observed data spacing, and further improvement in the subdivision accuracy affect slightly the inversion accuracy. Therefore, it is suggested that the model subdivision spacing for inversion could be half of the observed data spacing. The applications of the model tests and field magnetic data from a mining area demonstrated that the inverted magnetization obtained using the size of the model space subdivision of 0.5 times the observed data spacing are significantly better than those obtained using the general 1 time the observed data spacing.</p>