The Northern Territory (NT) in Australia has been perceived by many as a frontier for future agricultural development that could serve as a future food bowl for Australia. However, such development is likely to require the conversion of large areas of relatively intact savanna forests, woodlands and grasslands into other land uses, resulting in the release of large amounts of greenhouse gases. With the evolution of international carbon markets, new livelihood opportunities are arising from the management of carbon in landscapes, to enhance sequestration as a means of combating climate change. In order for land owners and governments to realise these opportunities, it is necessary to gain a better understanding of the potential carbon stocks that can occur across different ecosystems throughout the NT. This article assesses the utility of the National Carbon Accounting Toolbox (NCAT) for estimating and mapping carbon stocks in the NT. The NCAT modelling estimates that the NT environments hold approximately 21.5billion tonnes of carbon dioxide equivalents. The estimates of carbon vary from 5 to 235tonnes of carbon per ha, decreasing from the north, where the highest estimates are for north-east Arnhem Land and the Tiwi Islands, to the arid lands of the south. They are thought to be the lowest in the Spinifex grasslands east of Alice Springs and west of Tennant Creek. Estimates of potential emissions from clearing and burning of native vegetation range from 27 to 439tonnes of carbon dioxide per ha, depending on the initial vegetation type. The NCAT estimates are already being used for land-clearing assessments in the NT and demonstrate that Indigenous lands hold high levels of carbon. The performance of NCAT will depend strongly on the quality of the data on which predictions are based and the robustness of model parameterisation. We suggest that further work on soils, fire, grasslands, wetlands and woody debris is needed to improve the validity of the NCAT estimates for carbon in north Australian environments.