Understanding what drives, catalyzes or constraints land use change in the Brazilian agricultural frontier is a condition for effective policy design at the local level, which in turn might have implications for food production, environmental conservation and greenhouse gas emissions worldwide. We analyzed the process of agricultural expansion observed in the state of Mato Grosso, the country's largest agricultural producer, by mapping and quantifying the incorporation of new farming areas and the conversion of existing ones into mechanized soybean fields at the farm-level. Through statistical modelling we also investigated the influence of key economic, biophysical, environmental and logistics variables on this process while accounting for recent changes in the Brazilian environmental legislation. We found that the area converted to soybean production increased almost 1.5 million hectares between 2009 and 2013, more than 70% of which in farms that already had some soybean in previous years. By comparing the explanatory power of eight regression models involving different groups of variables, we found that soybean expansion is strongly associated with the presence of other soybean fields and warehouses within 50-100 km. The model with the largest explanatory power suggests that soybean expansion is also likely to occur in areas of high conservation value. Finally, the sensitivity of soybean expansion to soybean prices indicated the potential for further agricultural growth in Mato Grosso while highlighting how crucial smart logistics investments are for regional development with environmental protection.