Oribatid mites are an ancient group of cosmopolitan terrestrial arthropods with limited trans-oceanic dispersal abilities. They provide an opportunity to answer questions concerning the role played by Gondwanaland, either as a migration route for terrestrial organisms or as a centre for their origin and subsequent glacial destruction, in the development of the biota of Antarctica, the sub-Antarctic islands and nearby land areas. Biogeographical studies of present-day oribatid mite faunas of the Antarctic region, New Zealand and South America (particularly the Andes Mountains) also allow insight into the historical development of such biota after the break-up of Gondwanaland. No records of fossil oribatid mites are known for the Antarctic and their main dispersal mechanisms within the biome are likely to be via sea-birds and possibly ocean currents. A total of 105 species from 20 families of oribatid mites are recorded from the Antarctic which, together with species records from South America, including Patagonia and Tierra del Fuego, and New Zealand, allowed faunal similarities to be examined using the similarity coefficient of Jaccard and principal co-ordinate analysis. Species endemism is high in both the continental (60%) and the sub-Antarctic zones (63%) compared with the maritime Antarctic zone (18%) and the Falkland Islands (19%), but lower than in the New Zealand fauna (83%) and in the Neotropical areas of South America (89%). Species diversity of oribatid mites in the Antarctic is low (five species in the continental Antarctic zone, 22 species recorded for the maritime Antarctic zone, and 78 species found in the sub-Antarctic zone) compared with New Zealand (330 species) and the Neotropical South American region (1193 species). The numerically-dominant species are from the families Oppiidae and Ameronothridae in the Antarctic region, but only a single endemic family (Maudheimiidae) occurs there. Several conclusions are drawn regarding the relationships of the oribatid mite faunas within Antarctica and between them and those of the surrounding land areas. The high similarity of the present faunas of the Andes Mountains and New Zealand at both generic and family levels suggests a genetic continuity of these areas in the past, but reduced species similarity indicates that the majority of the present oribatid mite species arose after the break-up of Gondwanaland (17 oribatid mite species found in both areas at present, have not been recorded elsewhere). Two possibilities regarding the possible land connection between these two geographical areas are (i) via what are now the sub-Antarctic islands, and (ii) via what is now the continental Antarctic. The latter is more probable in view of the disharmonic nature of the present sub-Antarctic island oribatid mite faunas. The present distribution of the most common family of oribatid mites, the Ameronothridae, in the Antarctic suggests that it is not a faunal relict but results from post-glacial recolonisation possibly from Australasia, where there are many species and high generic diversity. The sub-Antarctic islands have a distinctly richer oribatid mite fauna than the continent, probably influenced by a longer period of habitat development uninterrupted by volcanic activity. Their re-colonisation by physiologically pre-adapted, cold-hardy taxa is likely to have occurred from South America and Australia, a long-term and often accidental process. Studies are required to determine the possible mechanisms which underlie this process.