Numerous unresolved issues exist regarding the Antarctic lithosphere, especially in terms of density, temperature, and compositional structure. For example, contemporary estimates of the Moho depth from different geophysical methods show significant discrepancies of 10-20 km in large areas of the continent. While seismological methods suffer from a limited station coverage and ice reverberation, potential field methods, such as gravity studies, are inherently non-unique. By modelling densities and seismic velocities in a consistent way and accounting for thermodynamically stable mineral phases of rocks under in-situ pressure and temperature conditions, we are able to compensate for the sparseness of data in Antarctica and reduce inconsistencies and ambiguities of separate geophysical methods to a large extent. Gravity gradient data from ESA’ s satellite mission ‘GOCE’ are used to constrain the density distribution within the lithosphere in an integrated 3D model of the Antarctic continent, whereas independent seismic estimates serve as a benchmark for the robustness of the results. Intra-crustal density variations are inverted from airborne gravity data and evaluated against thermodynamic modelling results of potential upper crustal rock compositions. As a main outcome, we present a new crustal thickness map of Antarctica. It includes varying density contrast at the Moho and exhibits detailed thickness changes in East Antarctica, also in regions with sparse or absent seismic station coverage. At locations where compositions of xenoliths from deeper sources are available, we analyse the model’s sensitivity to introduction of such regional information. As an example of the applicability of our model for further studies, we demonstrate how its deep thermal field is used to derive upper mantle viscosities for glacial isostatic adjustment (GIA) modelling. @en