A refined model of the late Mesoproterozoic to Neoproterozoic supercontinent Rodinia is presented, with Baltica, Amazonia and West Africa attached to eastern Laurentia as in the SAMBA model (Johansson, 2009), East Antarctica, Australia and India to western Laurentia in a SWEAT configuration (Moores, 1991), and northern Siberia attached to northern Laurentia as proposed by Condie and Rosen (1994). In such a model, the Proto-Andean margin of South America would form the conjugate margin of Laurentia’s Grenville margin. With the Kalahari craton attached to SW Laurentia and East Antarctica, as proposed by Loewy et al. (2011), followed by the Congo and Tanzania cratons in Africa and the Sao Fransisco and Rio de la Plata cratons in South America, all these cratons would be part of Rodinia, but would still be separated from Amazonia by a wide Brasiliano (Clymene) ocean embayment. By rotating the African and eastern South American cratons ca 90° counterclockwise around a pole located close to the Laurentia – Kalahari junction, and East Antarctica, Australia and India ca 120° counterclockwise around a pole located inside the Kalahari craton, all relative to a fixed Laurentia, these cratons will move from a Rodinia to a Gondwana configuration. These rotations will open up the Proto-Pacific ocean, close the Brasiliano (Clymene) ocean, and both open and close the intervening Adamastor and Mozambque oceans, creating the various Brasiliano and Pan-African fold belts in the ensuing collisions. The maximum plate velocity, ca 7.5 cm/year (15 000 km in 200 m.y.), will occur along the outer periphery of this rotation, thereby explaining the formation of large amounts of juvenile Neoproterozoic continental crust within the oceanic Arabian – Nubian sector of the Pan-African Orogen. Rather than being an example of ‘introversion’ or ‘extroversion’, the change from Rodinia to Gondwana in this model would be more like the 90º ‘orthoversion’ model proposed by Mitchell et al. (2012).