Measuring arrangements with variable light conditions show illumination-dependent influences as deviations of the measured target‘s centre, which is detectable in both image and object space. To analyse the causes, an experimental setup was realised for the experimental investigation of different lighting situations, where the illumination-dependent behaviour was confirmed: deviations up to 1 pixel were measured in image space under suboptimal conditions. Influencing factors for these deviations are the tar-get‘s material and its specific roughness profile as well as the beam geometry, depending on the direction of view. Known reflection models, for example in the field of computer graphics (Phong, Cook-Torrence, etc.), allow to compute beam geometry and reflection behaviour very well. Nevertheless, the models cannot conclusively explain the relocation of the target’s centre or provide a corrective approach. Therefore, experimental studies have been conducted, which show a systematic effect. In a first step, a correction of the target diameter could be achieved. In the second step, an optimized target design was developed, which both significantly increased the stability of the image measurement and, under certain conditions, reduces the illumination-dependent shift of the ellipse centre.