Despite decades of study the prerift configuration and early rifting history between Australia and Antarctica is not well established. The plate boundary system during the Cretaceous includes the evolving Kerguelen–Broken Ridge Large Igneous Province in the west as well as the conjugate passive and transform margin segments of the Australian and Antarctic continents. Previous rigid plate reconstruction models have highlighted the difficulty in satisfying all the available observations within a single coherent reconstruction history. We investigate a range of scenarios for the early rifting history of these plates by developing a deforming plate model for this conjugate margin pair. Potential field data are used to define the boundaries of stretched continental crust on a regional scale. Integrating crustal thickness along tectonic flow lines provides an estimate of the prerift location of the continental plate boundary. We then use the prerift plate boundary positions, along with additional constraints from geological structures and large igneous provinces within the same Australian and Antarctic plate system, to compute “full‐fit” poles of rotation for Australia relative to Antarctica. Our preferred model implies that the Leeuwin and Vincennes Fracture Zones are conjugate features within Gondwana, but that the direction of initial opening between Australia and Antarctica does not follow the orientation of these features; rather, the geometry of these features is likely related to the earlier rifting of India away from Australia‐Antarctica. Previous full‐fit reconstructions, based on qualitative estimates of continental margin overlaps, generally yield a tighter fit than our preferred reconstruction based on palinspastic margin restoration.