Divergence of the Australian and East Antarctic plates is well understood from the late Jurassic onset of half graben development on the Australian continental shelf, and from post mid-Eocene (chron 20; 45 Ma) seafloor spreading isochrons further offshore. Relative plate motion between these times is less confidently interpretable from magnetic reversal anomalies landwards of isochron 20 and localised evidence for mid-to-late Cretaceous subsidence and growth strata from the continental shelf and rise south of Australia. A new test of this history examines it within the post-34y (84 Ma) Indian Ocean plate circuit, built using seafloor spreading data from the Wharton and central Indian Ocean basins. The Australian-Antarctic Jurassic-onset rift system is interpreted to have been abandoned before 34y, because motion in the circuit is inconsistent with an active plate boundary during 34y–26y (84–58 Ma). Starting 26y, the model depicts plate divergence distances and azimuths that, after 25y (57 Ma), can be independently confirmed by reassessment of the pre-chron 20 magnetic reversal anomaly pattern. Previous studies have identified evidence for mantle exhumation and focused magmatism in basement marginwards of these anomalies. These processes are not directly or confidently dated, but mantle exhumation is inconsistent with the circuit model's fast plate divergence at 26y–25y. Hence, plate motion during the immediate build-up to post-57 Ma seafloor spreading may have been accommodated by focused magmatism, whilst mantle exhumation may mark the conclusion of the Jurassic-onset rift phase during a slower pre-84 Ma period of plate divergence. Using the new model to make tectonic reconstructions results in a large overlap between Tasmania and Victoria Land that can be explained with reference to Eocene strike-slip faulting and transtension in recently- discovered subglacial basins of Wilkes and George V lands and Terre Adélie.