It is widely acknowledged that in the terrestrial Antarctic, interspecific interactions are typically unimportant in determining species distributions and community structure. Therefore, correlative models should prove useful for predicting current and future spatial variation in species abundance patterns. However, this idea has not been formally tested, and the utility of such models, which have shown value for understanding the distribution of diversity elsewhere, for investigating biodiversity patterns in Antarctica remains unclear. Here we make a start at such tests by using generalized linear and simultaneous autoregressive models to demonstrate that simple environmental variables and information about the spatial structure of the environment can explain more than 90% of the variation in the abundance of Maudheimia wilsoni (Oribatida; Maudheimiidae), a representative of one of the most significant groups of Antarctic terrestrial arthropods, the mites. We show that a single environmental variable, maximum soil moisture content, can account for as much as 80% of the variance in the abundance of the mite, and that linear models with only a few environmental and spatial terms can be used to forecast the species abundance at the landscape scale. Given ongoing calls for better understanding of the distribution of Antarctic diversity and its likely future change, this initial test indicates that such modelling procedures, and more sophisticated versions thereof, hold much promise for the region and should be tested for other taxa with different life forms and habitat requirements.