The potential spread of any invasive plant is a central concern in weed risk assessment. Calotropis procera is wind dispersed and forms extensive monospecific stands that reduce the productivity of pastoral land, but its potential distribution and drivers of its spread are not well known. Using maximum entropy methodology, we modelled current and future potential distributions of C. procera in Australia. Occurrence data (n?5976 presence records) were collated from regional databases and a field survey. Of a set of 'independent' environmental correlates, those that best accounted for the observed distribution of C. procera in Australia were distance (km) to roads, average annual rainfall (mm), mean temperature (°C), average wind speed (km/h), beef density and vegetation type, in that order of importance. Current and potential distribution of C. procera was best explained by interactions between anthropogenic disturbance and climatic factors, all underpinned by species characteristics. Models were based on a grid cell size of 5km x km and model performance was good (mean AUC = 0.916; s.d. = 0.014; AUC = area under the curve; perfect fit = 1). The model showed that C. procera has not saturated its current potential distribution. Models of future spread derived from climate change projections, based on global circulation models in the 'Representative Concentration Pathway 4.5 emissions scenario for 2035', show the area suitable for C. procera will increase, increasing the risk the weed poses. Range expansion will occur into all three states surrounding the Northern Territory, but mostly into the north-eastern border regions of Western Australia and north-western Queensland. Joint management of rubber bush at a regional scale across jurisdictions, is urgently advised to avoid future spread of rubber bush and further reductions in pastoral productivity. � Australian Rangeland Society 2016.