Due to its excellent soft tissue contrast, magnetic resonance imaging (MRI) is widely used for target delineation in stereotactic radiosurgery. However, because of a number of factors (inhomogeneity of the constant magnetic field, nonlinearity of the gradient fields, etc.), MR images are more susceptible to spatial distortion as opposed to Computed tomography (CT) images, which could be considered as the reference studies in terms of the accuracy of target localization. It is crucial to assess the geometric errors in MR images as the accuracy of dose delivery strongly depends on MRI quality. One of the ways of MRI distortion evaluation is to scan a phantom with multiple marker points on both CT and MRI scanner and compare coordinates of the points defined on CT and MR images, respectively. Performing this test as a routine part of MRI quality assurance (QA) program for stereotactic radiosurgery is complicated by lack of a software for the automatic calculations of the discrepancies.The aim of our work was to develop a software to calculate discrepancies between stereotactic coordinates of the phantom’s plastic rods defined on the MR and CT images of the Elekta MR phantom. A special Python software was developed and the spatial accuracy of the stereotactic MR images was assessed. The calculated total differences between the coordinates of the CT and T1 weighted MR images, CT and T2 weighted MR images exceeded 1 mm in 3,5%, 0.1% of the points, respectively. Using the software in clinic routines could both speed up the time needed for the test performance and eliminate subjective visual assessment of the discrepancies.