Abstract. It has been assumed, that the maximum compressive horizontal stress (SHmax) orientation in the upper crust is governed on a regional scale by the same forces that drive plate motion. However, several regions are identified, where stress orientation deviates from the expected orientation due to plate boundary forces (first order stress sources), or the plate wide pattern. In some of this regions a gradual rotation of the SHmax orientation has been observed. Several second and third order stress sources have been identified, which may explain stress rotation in the upper crust. For example lateral heterogeneities in the crust, such as density, petrophysical or petrothermal properties and discontinuities, like faults are identified as potential candidates to cause lateral stress rotations. To investigate several of the candidates, generic geomechanical numerical models are utilized. These models consist of up to five different units, oriented by an angle of 60° to the direction of contraction. These units have variable elastic material properties, such as Young's modulus, Poisson ratio and density. Furthermore, the units can be separated by contact surfaces that allow them so slide along these faults, depending on a selected coefficient of friction. The model results indicate, that a density contrast or the variation of the Poisson's ratio alone sparsely rotates the horizontal stress orientation (≦ 17°). Conversely, a contrast of the Young's modulus allows significant stress rotations in the order of up to 78°; not only areas in the vicinity of the material transition are affected by that stress rotation. Stress rotation clearly decreases for the same stiffness contrast, when the units are separated by low friction discontinuities (19°). Low friction discontinuities in homogeneous models do not change the stress pattern at all, away from the fault; the stress pattern is nearly identical to a model without any active faults. This indicates that material contrasts are capable of producing significant stress rotation for larger areas in the crust. Active faults that separates such material contrasts have the opposite effect, they rather compensate stress rotations.