The drift history of Gondwana following the break-up of Rodinia (or perhaps Pannotia) to the amalgamation into Pangaea has great implications in many disciplines in Earth sciences, but remains largely unknown. Among the apparent polar wander (APW) paths published for Gondwana in the last few decades, large discrepancies exist (sometimes up to thousands of kilometres). The mid Palaeozoic segment of the APW path is particularly problematic, and two primary schools of thought arise. Some authors favour a Silurian – Devonian loop in their APW path passing through southern South America (on a reconstruction of Gondwana), whereas others draw a path directly through Africa during this period. The main controversy stems essentially from whether or not palaeomagnetic data from eastern Australia are incorporated in order to compensate for the lack of mid Palaeozoic data. Determining whether the terranes of the Southern Tasmanides are (para-)autochthonous or allochthonous in origin is therefore of crucial importance and a matter of intense debate. The aim of the work presented herein is to palaeomagnetically define the positions of these terranes throughout the Palaeozoic in order to better constrain the complex tectonic history of this region and to help clarifying the APW path of Gondwana. The construction of an APW path is discussed herein. An attempt is made to determine whether only “objective” criteria can be employed to select data used to draw an APW path. However, it is shown that the palaeomagnetic database has not enough entries. Subjective data selection must be introduced leading to two end-members: the X-type and the Y-type, thought to be best illustrated by the X-path proposed by Bachtadse & Briden (1991) and the Y-path proposed by Schmidt et al. (1990). These two models are, therefore, used in the discussion of the results obtained for this study. The Southern Tasmanides had a complex tectonic history with several orogenic events throughout the Palaeozoic. The sampling coverage carried out for this study comprises fifty localities (289 sites, 1576 cores, 3969 specimens; see table 1, pages 54-55) distributed along an east-west transect across most of the subdivisions of the Southern Tasmanides. The sampled localities are gathered in three main areas: the Broken Hill area, the Mount Bowen area, and the Molong area, which are situated where no published palaeomagnetic studies were previously available providing, therefore, new information. Sampling and laboratory procedures have been carried out using standard techniques. In particular, detailed stepwise thermal demagnetisation, principal component analysis, anisotropy of magnetic susceptibility and rock magnetic measurements have been systematically employed. The routine measurement of the anisotropy of magnetic susceptibility allowed drawing the first maps of the magnetic fabrics throughout the region. A strong correlation between the magnetic fabrics and the main tectonic structures corroborates the existence of cross-structures (E-W) in the Southern Tasmanides. The directions of magnetisation obtained yielded much information, despite poor quality. The effects of weathering are deep, intense and widespread. For example, most of the samples from the Mount Arrowsmith Formation (localities ARR & ARO) and the Funeral Creek Limestone (FUN) in the Broken Hill area (western New South Wales) are totally remagnetised, as well as some from the Mitchell Formation (MIT) in the Molong area (eastern New South Wales). Secondary magnetisations are also largely responsible for the bad results obtained in most of the fifty localities studied. Intermediate directions of magnetisation are common and often result in significant data scattering, as illustrated for instance by results from the Kandie Tank Limestone (KAN; Broken Hill area) or the Ambone and Ural Volcanics (HOP, BOW, SHE; Mount Bowen area). In general, it has not been possible to precise the remagnetisation process leading to th