A consolidated knowledge of the formation and dispersal of the former supercontinents reveals important evidence for the earth’s climate and biosphere in the past and contribute to the prediction of their future evolution. Nowadays, a main objective is the investigation of the initial break-up of the continental assembly of Gondwana that serves as a constraint for its subsequent dispersal and the evolution of all oceans and seas in the southern hemisphere. Evidence of the early rifting stages are expected at the margins of Southeast Africa and East Antarctica, whereas the latter one is difficult to access, due to its remote position and ice coverage. To understand the driving forces and the chronology of the break-up and of the massive volcanism additional detailed knowledge of the crustal setting along the margins of Southeast Africa is required. Therefor, a new geophysical dataset was acquired with the RV Sonne in the northern Mozambique Basin at the beginning of year 2014. This comprises a deep seismic sounding profile across a so-far unknown structural high, the Beira High. Additional gravity and magnetic data were systematically recorded across the entire northern Mozambique Basin. Based on velocity, amplitude, density and magnetic modelling, a geological model of the continental margin of Central Mozambique was prepared. A new compilation of all available magnetic data in the Mozambique Basin reveals information about the age of the sea floor, which serves as constraint for the reconstruction of the initial Gondwana break-up. The study depicts a continental origin of up to 23 km thick and partly highly intruded crust at Beira High. In the adjacent coastal areas of the south-western part of Central Mozambique, 7 km thin crust is observed, which is covered by more than 11 km thick sediments and implies the continuation of the continent-ocean transition towards onshore Mozambique. This is in clear contrast to the narrow transition observed in the north-eastern part of the margin and reveals a clear asymmetric crustal setting, as supposed for the conjugate margin in the Riiser-Larsen Sea in Antarctica and consequently suggests a complex break-up scenario. The presence of a pronounced high-velocity lower crustal body is interpreted as magmatic material, which underplates the crust and extends about 200 km from the Central Mozambican margin towards the Mozambique Basin and testifies for the massive volcanism during the break-up. The distribution of further volcanics along the entire margin clearly depicts the continuation of the north-eastern branch of the Karoo large igneous province and are mainly emplaced between 177-157 Ma. The magmatism in Southeast Africa seemed to be continuous throughout the initial break-up, which points to the presence of either a mantle plume or a thermal anomaly as source of the giant magmatism. An additional late stage of rift-volcanism mainly affected the margin of Dronning Maud Land and causes the difference in the magnetic signature of the conjugate margins. The tracing of continuous fractures throughout the Africa-Antarctica Corridor leads to the reconstruction of a tight Gondwana fit prior rifting, which reveals several geological links between the plates. A main structure of the East African-Antarctic Orogen extends from the Namama Shear Zone in Central Mozambique across the Orvin Shear Zone towards the Forster magnetic anomaly in Dronning Maud Land. During the initial Gondwana break-up at 182 Ma, Beira High started to separate from West Gondwana along this suture until it demerged as well from East Gondwana by a rift jump. The investigation of further partly unknown tectonic structures along the western and southern coast of Mozambique revealed a possible oceanic origin of the southern part of the Mozambique Coastal Plains, due to similarities of the magnetic signature to the oceanic crust south of Beira High as well as the tentative identification of magnetic spreading anomalies. The subsequent emplaced Mozambique Ridge moved southwards as part of a micro plate during an additional active spreading centre in the Northern Natal Valley. The resulting reconstruction of the initial Gondwana break-up in the Africa-Antarctica Corridor accounts for all present-day available geological, geophysical and geodynamic constraints and might serve as a basis for the investigation of the subsequent dispersal of Gondwana.