The accurate simulation of stray light is essential for the verification of the contrast requirements in optical instruments. In a spectrometer, the scattering from reflective gratings is difficult to characterize while contributing significantly to the overall system stray light and reduction of the spectrometer contrast.Curved gratings also introduce higher straylight and difficulties in the verification of their straylight compliance. Scatterometers have to be adapted to the grating curvature and thus loose some of their performance.While measuring a curved sample, this becomes part of the optical system of the scatterometer. For the AlbatrossTT and the CASI the system considering the curved sample has to be adapted to refocus the beam onto the detector. Additional aberrations modify the near angle limit, i.e. the smallest angle where the measured BRDF is higher than the system signature. Theoretical analysis of the scatterometer becomes necessary because the signature of the system is no longer directly measureable.In addition, 2D BRDF measurement is adequate for characterization of a high polished optical surface or black coatings but deems to be insufficient to characterize all features of the grating scatter function.3D measurement is essential to verify compliancy in all angular directions.Small errors in the measurement procedure or measurement setup can lead to different results. Also every scatterometer setup has its own design driven limitations effecting near angle limit and resolution. To create reasonable results those limitations have to be carefully considered.We present a set of measurements in two different laboratories with different scatterometers.