A growing number of mines in Queensland, Australia, are moving towards rehabilitation certification and relinquishment as a result of the changing economic, social and environmental climate. While the mining industry aspires to the rehabilitation goals of safe, stable, sustainable and non-polluting, this aspiration is hampered by a lack of knowledge on rehabilitation resilience and recovery from inevitable future disturbances such as fire and drought. The aim of this study was to apply an experimental fire to established rehabilitation at the Curragh Mine in Central Queensland and determine the vegetation response and resilience in the two years following the burn. The project used ecological ground-based and remote sensing assessments to address the research questions that quantified fire behaviour, fire severity and vegetation recovery trends to inform stakeholders of short-and long-term opportunities and threats from fire impacts on rehabilitated lands.The first research question considered fire behaviour on established (19-to 21-year old) rehabilitation, and related ground fuel loads to a number of fire behaviour metrics including rate of spread, fire intensity and fire severity. The site was stratified based on two common site preparation techniques employed at the Curragh Mine site: i) the complete surface coverage (100 %) of topsoil that developed into a grassland community and ii) topsoil applied across the surface in 10 metre wide strips aligned with contours, that resulted in an open woodland community. Results of the pre-fire surveys showed that grassland areas had significantly higher ground fuel loads (p<0.0001) and as a result, burnt at a higher intensity (p<0.05) compared to open woodland transects. Due to a continuity of ground fuel loads, fire rate of spread was higher in grassland areas and fire severity metrics showed that grassland areas burnt at a higher severity compared with open woodland areas. Our results indicate that within the range of study conditions, rehabilitation technique has a significant effect on fire behaviour and therefore fire risk at this site, particularly when ground fuel loads exceed 13.5 t/ha. The second research question studied fire severity and recovery using remote sensing techniques and classified imagery captured by Unmanned Aerial Vehicle (UAV) and WorldView-3 satellite in a time-series throughout the course of the project. A number of vegetation and fire indices were tested to determine the effectiveness of categorising fire severity and vegetation recovery post-fire. Although UAV imagery was limited to indices Page 3 of 197utilising 3-band red, green, blue (RGB) products, the differenced Excess Green Index (dEGI) was shown to satisfactorily categorise fire severity impacts across the site, with an overall map accuracy of 72 % and kappa 0.55. Compared with UAV imagery, the WorldView-3 sensor provided a precise and accurate high resolution data source in all dimensions; with eight-band multispectral (1.3 m pixel) and eight-band short wave infra-red (7.2 m pixel)...