The Central-Eastern Mediterranean region is known to be a complex area due to the interaction of four tectonic plates, namely Arabia, Africa, Anatolia and Eurasia, and by the presence of an ancient oceanic crust in the Herodotus and Ionian Basins. The analysis of the available literature highlights that the distribution of the freely available geophysical data (i.e. seismic, gravity and magnetic observations) is quite disparate. In this framework, high-resolution global gravity field models, such as XGM2019e, mainly based (offshore) on satellite data, can be profitably used as a uniform dataset to study, in a coherent way, large regions.In the current work, we exploit the XGM2019e model, together with a set of a priori information, derived mainly from geophysical data retrieved from the literature, to study the structure of the crust in the Central-Eastern Mediterranean area. The study is organized in three different phases: in the first one, we enhanced the map of geological crustal provinces by means of an automatic Bayesian classification algorithm applied to the second radial derivatives of the gravitational potential. In the second phase, using as observation a grid of gravity anomalies, we applied a full three-dimensional inversion procedure (always based on a Bayesian paradigm) to estimate the mass density variations and the geometries of the main geological units in the whole study area. Finally, in the third phase, we performed a refined three-dimensional local inversion on the Cyprus area to improve the modelling of the shallowest layers. The main results of this study, carried out in the framework of the European Space Agency GIADA project, are freely available, upon request, at https://www.g-red.eu/geophysics/.