In the classical XRD residual stress methods omega-stress and psi-stress, an intensity correction for absorption is commonly applied. A related effect is transparency, which causes asymmetric peak broadening and peak shifts. Since the average depth of information is below the sample surface, this implies a specimen displacement effect, if the sample surface is used as reference for mounting. Mounting the sample with its information depth coinciding with the tilt axis doesn't help due to the varying values for different tilts. For most metals the peak shift effect due to transparency can be safely neglected in combination with Cu, Co or Cr radiation. For X-ray transparent samples (e.g. ceramics or metals with Mo radiation) a peak shift correction will be necessary. This paper proposes a peak shift correction due to transparency for the general case of a partly transparent layer. Additionally two extreme cases are discussed: bulk samples and fully transparent layers.