In earth’s history, large assemblies of continental masses, so-called supercontinents, occurred episodically. The last ones, Pangaea and Gondwana, broke up in Triassic to Jurassic times. The distribution of the landmasses has major influences on the climate and biosphere on earth. To have detailed knowledge about the timing and devolution of the assembling and decomposition processes of these large continents is thus important as framework for many earth-related scientific disciplines. This study deals with the movements between Africa and Antarctica in Middle Jurassic to Lower Cretaceous times. Only a thin stripe of seafloor, the Africa-Antarctica corridor, provides direct evidence of these movements. Although the kinematic history between the two continents is quite well constrained for Cenozoic times, little is known about the Jurassic processes, mainly due to the lack of data on the conjugate continental margins. New wide-angle seismic and potential field data were thus acquired on both conjugate margins during four scientific campaigns between 2006 and 2010 by AWI and co-operation partners to gather information about the age and origin of the crust in the Africa-Antarctica corridor. On both conjugate sides, aseismic plateaus exist, whose origin is unclear. The assumption of continental crust underneath the Mozambique Ridge, the Northern Natal Valley and the Mozambique Coastal Plains produces significant overlaps with Antarctica in many Gondwana reconstructions. On the conjugate margin, the Astrid Ridge is a poorly studied feature whose history is apparently related to the formation of the oldest crust in the Africa-Antarctica corridor. Aeromagnetic measurements in the southwestern Enderby Basin/Cosmonauts Sea east of the Gunnerus Ridge were conducted in 2006. The data clearly image the continent-ocean-transition offshore Prince-Harald Coast. No magnetic spreading anomalies can be seen in the data, however. The oceanic crust is thus interpreted of having formed during times of the Cretaceous Normal superchron. India and Sri Lanka were thus attached to Antarctica up to these times. Two wide-angle seismic profiles across the Central Mozambique continental margin as well as magnetic and gravity data were acquired in the Mozambique Channel in 2007. P-wave velocity and density modelling of the seismic refraction profiles reveals continental crust, thinning seawards by 50% over a distance of 130 km. From a pronounced negative magnetic anomaly near the shoreline southwards, slightly thickened to normal oceanic crust is present. Lowermost sediments with high P-wave velocities around 1 Abstract 4.8 km/s overlie the oceanic basement. An extensive lower crustal high-velocity-body with P-wave-velocities 7.0 km/s has been found in both profiles. The identifications of magnetic spreading anomalies in the Mozambique Channel have been extended to the north, revealing that the continent-ocean-transition is located closer to the coast than supposed so far. M41n is interpreted to be the oldest existing magnetic spreading anomaly near the Central Mozambique shoreline. Extensive and systematic magnetic and gravity measurements have been made on the Mozambique Ridge and the Natal Valley in 2009. The data show magnetic spreading anomalies with different trends on the Mozambique Ridge and in the Northern Natal Valley, which are thus mainly underlain by oceanic crust. The similarity of the magnetic and gravity field between the Mozambique Coastal Plains and the new data farther south and the absence of signs for a continent-ocean-boundary give evidence for a mainly oceanic nature of the Mozambique Coastal Plains as well. In 2009/2010, systematic aeromagnetic measurements were made across the Astrid Ridge and the southwestern Riiser-Larsen Sea, Antarctica. The Astrid Ridge is subdivided by the Astrid Fracture Zone into a northern and a southern part with different magnetic signatures. The southern part is weakly magnetized as is the southwestern Riiser-Larsen Sea. It is interpreted to consist of oceanic crust. Parallel anomalies on this part of ridge point to a formation by seafloor spreading in another direction than the Riiser-Larsen Sea. The northern part of the Astrid Ridge shows strong positiv magnetic anomalies, pointing to a different and younger formation history than the southern part. The regional results were interpreted within the context of Gondwana breakup and implemented into a new kinematic model between Africa and Antarctica, which is presented in this study. This model adapts a tight continental fit for Gondwana and postulates a two-stage breakup. During the first stage, Antarctica rotated anticlockwise with respect to Africa, the Grunehogna Craton clearing the Mozambique Coastal Plains and occupying a position east of the Mozambique Fracture Zone around 168 Ma. Afterwards, Antarctica moved southwards during stage 2 and oceanic crust in the Mozambique Channel formed. The Mozambique Ridge and the Northern Natal Valley were formed by seafloor spreading on a spreading centre that changed its position two times by jumps to the south. The new kinematic model is in agreement with the known geological and geophysical findings and explains the formation of the Jurassic and Early Cretaceous crust in the Africa-Antarctica Corridor without contradictions and overlapping problems.