We investigate effects of realistic Earth's lithospheric structures on measurements of SKS seismic waves birefringence, which could be perturbed by different factors. We present SKS measurements recorded in four different tectonic settings. For each case, a realistic lithospheric structure is assumed and synthetic SKS splitting measurements are compared with the field observations. Our results show that (a) in a simple case, where anisotropy is aligned in both the lithospheric mantle and the asthenosphere, the SKS measurements can be safely interpreted as dominantly related to asthenospheric mantle flow; (b) in case of multilayer anisotropy in the lithospheric mantle, SKS measurement can correspond to a combination of the different fast-axis orientations and intensities of the anisotropic layers, dominated by the layer with stronger anisotropy; (c) across orogens, where highly anisotropic (≥10%) crustal sections are present, a relevant percent (30–40%) of SKS measurement can be explained by crustal contributions, with additional challenge related to the different direction of retrieved and expected symmetry axis; and finally, (d) in subduction zones, even in the absence of mantle corner flow, subducted crust materials can interplay with the overriding plate to generate an interpretable SKS observation. We conclude that complex crustal and/or lithospheric anisotropy can lead to erroneous SKS splitting parameters interpretations in terms of mantle deformations. We show in particular that complex crustal anisotropy can produce both important time delay and back azimuthal pattern. We also confirm that SKS splitting parameters do not allow to identify the presence of more than two anisotropic layers in the mantle.