The tectonic evolution of the South Atlantic region including the Weddell Sea between the East Antarctic craton and the Antarctic Peninsula is crucial in understanding the processes involved in early Gondwana break-up during Mesozoic times.A large aeromagnetic dataset, acquired by the Alfred Wegener Institute during the EMAGE (East Antarctic Margin Aeromagnetic and Gravity Experiment) project in the years 1996 2002, gives new unequivocal constraints on the timing and geometry of the early opening between South America, Africa and East Antarctica.With the help of this high resolution magnetic anomaly data a complete set of Mesozoic seafloor spreading anomalies off western Dronning Maud Land (DML) and in the Riiser-Larsen-Sea (RLS) could be identified and dated with great accuracy.In the eastern Weddell Sea the succession of recovered magnetic anomalies starts in the north with anomaly C33n (74 Ma) and proceeds to the south until the Mesozoic M13n (137 Ma) anomaly. Further to the southwest along the coast of western DML progressively younger anomalies can be found and are interpreted as M17n and M19n (142 Ma and 146 Ma). Based on these identifications in the eastern Weddell Sea, known magnetic lineations in the Central Weddell Sea could be dated more reliably than before and lead to a consistent pattern of magnetic anomalies from 40 W to 10 E.As a result of this new time and geometric constraints on the development of the ocean floor in the eastern and central part of the Weddell Sea and the Riiser-Larsen Sea a plate tectonic model for the early break-up of Gondwana has been developed. According to anomaly M24 (155 Ma) found as the oldest seafloor spreading anomaly in the RLS and M22 (152 Ma) in the Somali Basin, these where the places were the first oceanic crust formed at the very beginning of the dispersal of Gondwana. This is followed by the opening in the Weddell Sea along a rift propagating from the south western corner of the Weddell Sea to the west coast of DML where the oldest anomalies in the EMAGE dataset are interpreted as M17n to M19n (142 Ma to 146 Ma). It was not before M12 (135 Ma) that the passage between Africa and East Antarctica was fully established, which connected the two oceanic basins in the Weddell Sea and the Riiser-Larsen Sea. Short after the opening of this seaway the final separation between South America and Africa began in the Cape Basin and the southern Natal Basin.The position of the Antarctic Peninsula within the early break-up process is very poorly documented in the geological and geophysical record. No ocean floor off the peninsula so far has been found which preserved any information on its position during and/or just before the initial steps of dispersal. Based on geophysical, geological and palaeomagnetic data, several models for the movements of the Antarctic Peninsula have been published in the last years. Here we will review a few of them and discuss them in the light of the new model for the opening of the Weddell Sea and the South Atlantic Ocean.