New gravity field models incorporate GRACE data as the best available data source for the low to medium wavelength but validation is difficult since comparisons with existing GRACE models will always be biased. Maybe the best independent data set on a global scale is therefore the CHAMP data. It is known that the accuracy of the CHAMP solutions is approximately one order of magnitude worse than the one of GRACE-only solutions. On the other hand and considering e.g. the degree difference RMS between PGM2007A and GGM02S, also dis-crepencies between these models occur which cannot solely be explained by numerical inaccuracies. Consequently, it has been investigated if the CHAMP solutions can serve as an indicator. This research shows results of the evaluation of the preliminary gravity field model PGM2007A with a CHAMP solution derived from two years of kinematic orbits. The quality of the global CHAMP solutions is further improved by a local refinement with Slepian functions which can make better use of the information in areas of high data density, e.g. in high-latitude areas. However, it is concluded that despite the data density the poorer quality of the CHAMP data is preventing a definite assessment of the quality of the PGM2007A.