In this thesis I use a predictive modelling framework to explore the use of space at nineteenth and early twentieth century Australian sawmills. Sawmills were a key component of early European settlement in heavily forested areas and are often associated with the development of significant infrastructure, such as roads and rail and sea transport networks. Despite their importance and potential for enhancing our understanding of early European communities, few studies have been undertaken on historical sawmills in Australia, particularly in relation to spatial organisation on a comparative level. A dataset of 20 nineteenth and early twentieth century sawmills was analysed and sawmills were found to fall into one of four main types ranging from small scale temporary establishments (Type A) to large, permanent sawmills with multi-faceted settlements and permanent infrastructure and support services (Type D). Analysis also revealed that sawmill features were spatially organised into industrial, intermediate and domestic zones. The model is applied to a case study, the Mill Point sawmill in southeast Queensland and results suggest a general validity of the predictive model and point to directions for further refinement and development. The study has implications for future studies of early industrial enterprises in Australia.