Finding the best fit for East- and West-Gondwana requires a detailed knowledge of the initial Jurassic movements between Africa and Antarctica. This study presents results of systematic and densely spaced aeromagnetic measurements, which have been conducted in 2009/2010 across the Astrid Ridge (Antarctica) and in the western Riiser-Larsen Sea to provide constraints for the early seafloor spreading history between both continents. The data reveal different magnetic signatures of the northern and southern parts of the Astrid Ridge, which are separated by the Astrid Fracture Zone. The southern part is weakly magnetised, corresponding to the low amplitude anomaly field of the southwestern Riiser-Larsen Sea. The northern Astrid Ridge bears strong positive anomalies. Several sets of trends are visible in the data. In the Mozambique Channel, we extended the existing magnetic spreading anomaly identifications close to the Mozambique margin. Based on these and on spreading anomalies in the conjugate Riiser-Larsen Sea, we established a new model of the early relative movements of Africa and Antarctica in Jurassic times, and introduce a detailed model for the emplacement of the Mozambique Ridge. The model postulates a tight fit between Africa and Antarctica and two stages of breakup, the first of which lasting until ~159 Ma (M33n). During this stage, Antarctica rotated anticlockwise with respect to Africa. The Grunehogna Craton cleared the Coastal Plains of Mozambique and occupied a position east of the Mozambique Fracture Zone. The southern Astrid Ridge is interpreted to consist of oceanic crust that was formed prior to the Riiser-Larsen Sea during this first stage. During the second stage, Antarctica moved southward with respect to Africa forming the Mozambique Basin and the conjugate Riiser-Larsen Sea. The Mozambique Ridge and the Northern Natal Valley were formed at different spreading centers being active subsequently.