This study investigates the ecology of one of the best known Australian marsupials, the Common Brushtail Possum Trichosurus vulpecula, in central Australia. Trichosurus vulpecula is one of few medium-sized mammal species that persist in arid Australia today. Its distribution within the arid zone has declined markedly since European settlement. Two populations, one within the East MacDonnell Ranges along the Hale River and the other on Irving Creek, a River Red Gum creek in the Petermann Ranges, were studied in the southern Northern Territory. Others locations in the region were visited opportunistically. Trie central Australian Trichosurus is not distinct genetically from populations elsewhere in Australia. The diet of T. vulpecula consisted of a range of leaves, flowers and fruits of perennial dicotyledonous species as well as some ephemeral herbs. Grasses were absent from the diet. Variation in the diet reflected seasonal availability in flowers and fruits. The species preferentially consumed at each site had significantly higher moisture content and dry matter digestibility than species not consumed. Preferred species included Amyema maidenii leaves (a mistletoe), Acacia spp. flowers and fruits, Santalum lanceolatum leaves (a shrub), Marsdenia australis leaves (a vine), Solarium quadriloculatum fruit (shrub) and Euphorbia spp. leaves (herb). Small amounts of invertebrate material were consumed throughout the year. Other non-plant material consumed included honeycomb and unfledged birds eg. Budgerigars. There were no significant differences in the diet between the sexes. Trichosurus vulpecula were found in six main habitats: Acacia aneura/Callitris glaucophylla on rocky hills; E. camaldulensis sandy creek-lines; mixed Acacia rocky hills, Rocky Eucalyptus creek-lines; Degraded drainage lines; and Wet gullies. Logistic regression modelling revealed a significant correlation between mistletoe species richness, higher levels of soil nitrogen and the presence of T. vulpecula. In habitats occupied by T. vulpecula species richness of mistletoes was associated with the absence of fire and the presence of reliable ground water supplies. Trichosurus vulpecula were highly mobile with mean home ranges at Hale River of 44.21 � 22.76 ha and considerably higher than those recorded in previous studies in Australia. Mean home ranges at Irving Creek were much smaller, at 4.99 � 1.46 ha and VII similar to that recorded in other studies in Australia. At both sites, males had larger home ranges and there was a high degree of overlap with other males and females. At the Hale River study site, T. vulpecula predominantly denned in caves or cavities in rocks, whereas at Irving Creek all den sites were in large Eucalyptus camaldulensis on the drainage line. Adult and pouch young sex ratios were at parity. During this study, T. vulpecula was found to breed continuously, with births recorded in almost all months. Growth of the young were more rapid than previously recorded for Trichosurus in Australia. This is interpreted as an adaptation for living in an arid environment, enabling the young to achieve independence before quality food supplies diminish. No single exotic predator or competitor was solely responsible for the decline of T. vulpecula in arid Australia, implying an interactive impact. Prey switching by dingoes from rabbits to T. vulpecula, macropods and echidnas followed the crash of rabbit populations at Hale River. Predation by dingoes on T. vulpecula was only recorded once, at the Irving Creek study site, where numbers of rabbits remained stable throughout the study. The impact of exotic herbivores occurred through habitat degradation rather than competition. Evaluation of the ecological data collected during this study generally supports current models of decline and extinction in medium-sized mammals in arid Australia, integrating the effects of predators, competitors, drought and fire. However, the importance of each factor on populations of T. vulpecula was found to vary depending on their location in the landscape. This study suggests two separate models to explain the decline of T. vulpecula in arid Australia after the arrival of Europeans. The first operates in the riparian lowlands and the second on the rocky ranges. In both models, prior to European settlement, T. vulpecula occupied refuge habitats characterised by readily available moisture for plant growth (run on areas and/or shallow water tables) and soils with higher soil nutrient concentrations. The impact of fires on these refugia was minimal, as Aboriginal burning practices protected them with mosaic burning generally preventing large-scale fires from developing. Following European settlement, the forces impacting on populations were different in the riparian lowlands from those affecting rocky ranges. In the riparian lowlands, the effects of rabbits and livestock together with predation were found to have the major impact on T. vulpecula populations. Fire was not a significant factor in these areas. In the rocky ranges, fire was the most significant factor affecting T. vulpecula populations. Introduced herbivores did not degrade these habitats as they did in the riparian lowlands because the rugged and steep nature of the ranges acted as a physical barrier. Similarly, predator numbers were lower because of the relative difficulty in moving over rough ground and the generally lower relative abundance of preferred prey such as rabbits. An adaptive management strategy needs to be implemented to determine the effects of different management regimes on T. vulpecula population viability. The key elements of a management strategy in the riparian lowlands involves the manipulation and monitoring of predators, rabbits and livestock numbers. In the rocky ranges, the key management strategy involves the implementation of a patch burning to prevent fires entering habitats occupied by T. vulpecula. Importantly, any management strategies should involve Aboriginal people. Trichosurus vulpecula is an important part of Aboriginal culture. Its decline is of great concern to many people and several of the remaining populations and potential reintroduction locations are on Aboriginal land. Because of their relationship with the land and the animals, people have both the knowledge of the animal and the skills (such as patch burning) to provide information to managers which will assist with management. To achieve these management directions a coordinated national education programme is required to inform and convince the Australian community that conservation of T. vulpecula is deserving of attention in arid and semi-arid Australia. This is particularly important given the perception that T. vulpecula is a common species throughout Australia, despite its massive decline in arid Australia since European settlement.