Although the development of passive margins has been extensively studied over a number of decades, significant questions remain on how mantle and crustal dynamics interact to generate the observed margin geometries. Here, we investigate the Orange Basin, located on the south-west African continental margin. The basin fill is considered to comprise a classic rift-drift passive margin sequence recording the break-up of Gondwana and subsequent opening of the South Atlantic Ocean. Based on interpreted seismic reflection data, a 3D geological model was first constructed. Subsequently, an isostatic calculation (Airy′s model) using a homogeneous middle and lower crust was applied to this geological model to determine the position of the Moho for an isostatically balanced system. Isostatic sensitivity tests were applied to the model and their gravity response was validated against different crustal structures for the basin. The best-fit model requires dense, presumably mafic material, in the middle and lower crust beneath the basin and an abrupt change to less dense material near the coast to reproduce the observed gravity field.