A big challenge for northern Australia is developing socially, culturally and environmentally sustainable land management outcomes. This is particularly the case for indigenous people who constitute the majority of the remote rural population and who hold much of the land – either as freehold, leasehold or under emerging Native Title arrangements. Most of these lands are relatively pastorally unproductive. However, there is much scope in fire-prone northern Australia for using fire management to develop market-based environmental services enterprises to generate additional income from the land. This includes carbon farming opportunities associated with savanna burning emissions abatement and carbon sequestration projects. Northern Australia is the most fire-prone part of a very fire-prone continent. On average, depending on seasons, around 250,000–450,000 km2 gets burnt in northern Australia each year – an area up to seven times the size of Tasmania. By comparison, in recent times, the largest fires in southern Australia burnt somewhere between 20,000 and 30,000 km2 – the catastrophic fires of 2002–2003 in south-eastern Australia. Much of the north is burnt during the late dry season – that is after the cooler months in the middle of the year, from August onwards. This burning takes place under fairly severe conditions of high temperature, low humidity and variable winds. Of the 430,000 km2 higher rainfall savanna region which receives an annual average of more than 1,000 mm rainfall, currently over half is burnt on average each year and two-thirds of this is burnt in the late dry season, particularly in western Cape York (Fig. 1). The proportion annually burnt in the lower rainfall region further south is generally <30%. An associated staggering statistic is that most of the country is burnt by fires exceeding 1,000 km2. We know, however, that biota including small animals and poorly dispersed seed-regenerating plants cannot survive such frequent, intense and extensive fires. For ecosystem management and improved carbon sequestration outcomes, it is obviously more desirable to have smaller, patchy early dry season fires to provide greater habitat heterogeneity – and if such burning is strategically placed (e.g. burning from roadsides and creek lines), it can help create barriers which assist in reducing the spread and impact of late season wildfires. There is ample direct and inferential evidence that this type of systematic fire management practice was regionally widespread prior to European settlement. Over many years, a number of indigenous managers, agency personnel and research colleagues in northern Australia have been looking for economic solutions to try to address this fire management problem by turning it into an opportunity. Since the late 1990s, that work has particularly focused on developing and applying greenhouse gas emissions abatement methodologies. The simple rationale behind these methods is that if you reduce the extent of relatively intense late dry season fires through more strategic fire management, you can substantially reduce the amount of fire emissions. Over time, this promotes greater incorporation of carbon into living (especially trees) and nonliving (litter, woody debris) biomass pools. Interestingly, there is no direct evidence to suggest that reducing fire impacts, or even eliminating fires over decadal scales, results in any increase in soil organic matter in these highly seasonal savanna systems. Under international greenhouse accounting rules, we only account for the relatively long-lived greenhouse gases methane and nitrous oxide. On that basis, savanna fires contribute between 2% and 4% (depending on annual fire extent) of Australia's national greenhouse inventory accounts each year. If we were allowed to account for carbon dioxide (CO2) emissions (the major product of burning), it is likely that savanna burning would contribute as much as a third of the national inventory. The reason that CO2 emissions are not accounted for is based on the assumption that the carbon that gets emitted in one burning season is incorporated in regenerating vegetation in the following growing season. However, such an assumption ignores the facts that there may be many months between burning and regrowth, and under recurring severe fires, biomass carbon stocks may actually be in decline. We totally underestimate the global significance of emissions from north Australian savanna fires. After many years of development, in 2012, a savanna burning emissions abatement methodology was formally approved under the Australian Government's agricultural offset scheme, the Carbon Farming Initiative, paving the way for approved projects to generate carbon credits. Since that time, we have extended the original emissions abatement methodology to incorporate most of the savannas receiving more than 600 mm (Fig. 2), and have continued to work on developing associated sequestration methods. A method accounting for sequestration into the nonliving woody debris biomass pool is on track to be formally approved in 2016, and a further method accounting for sequestration into living tree biomass is under advanced development but still some years away. A recently published assessment indicates that, collectively, these methods will generate substantial additional carbon credits in fire-prone savanna systems – in fact, dwarfing economic returns from pastoralism in many parts of the north. We are thus talking about a totally new land sector industry. Most of the initial developmental work came out of the Western Arnhem Land Fire Abatement (WALFA) project. WALFA was initiated at the behest of traditional owners in 1996, essentially as a landscape-scale fire management programme, but also with the intention of reconnecting people with country which had been mostly vacated by the 1950s. Early years involved substantial planning and capacity building activities, although the limited resources available continued to have limited impact on what was essentially a wildfire regime. From 2000, WALFA also incorporated a major scientific effort to account for and examine how reimposition of traditional forms of fire management could help significantly reduce greenhouse emissions. By 2004, the operational capacity of indigenous land owners and regional ranger groups, and the scientific basis for emissions accounting, was sufficiently advanced to attract the attention of ConocoPhillips, a multinational energy corporation, which was looking for a viable greenhouse gas and environmental project to offset development of its liquefied natural gas plant in Darwin Harbour. A 17-year contractual arrangement commenced in 2006. That arrangement requires annual delivery of 100,000 t.CO2-e (each t.CO2-e equates to one carbon credit), and enhanced environmental management, through the delivery of a fire management programme focused on strategic early dry season prescribed burning over the 28,000 km2 WALFA project area. The WALFA region includes two internationally significant fire-vulnerable habitats – (i) monsoon rainforest dominated by the regional endemic tree, Allosyncarpia ternata, which constitutes the largest area of monsoon rainforest in the Northern Territory, and (ii) endemic sandstone heath communities which are listed as nationally endangered due to regional-scale fire impacts. The results from WALFA have been stunning. Since the commencement of the offset arrangement in 2006, WALFA partners have delivered substantially in excess of their contracted emissions abatement requirement. This has been achieved by turning around the pre-WALFA fire regime from one that was mostly wildfire-dominated to one where most burning now occurs in the milder early dry season period. In fact, a recent assessment shows that the fire management programme delivered through WALFA is now far superior to that currently being delivered in three major regional conservation reserves, including the contiguous World Heritage Kakadu National Park (Fig. 3). More broadly, similar commercial emissions abatement fire management projects are springing up across the north. At the start of 2015, there were 34 registered fire and carbon projects across northern Australia, all in the higher rainfall area. Many more have since been registered or are under development, including in lower rainfall regions. These projects are not confined to indigenously held lands, but increasingly many pastoral enterprises are involved. Importantly, emissions abatement projects can be undertaken on lands under all tenure types, including National Parks and Defence properties. With ever-diminishing operational budgets for regional conservation reserves, implementing savanna burning projects contracted to local indigenous land owners and ranger groups is an obvious (but as yet untried) solution. From a situation where fire has, and continues to be, a recurrent problem across many parts of northern Australia, we are now seeing an opportunity where land owners and managers can take advantage of market-based incentives to deliver improved fire and environmental management outcomes. In a real sense, these regional fire and carbon economy initiatives light the way for broader discussion and development of a variety of market-based ecosystem services opportunities. Examples include developing additional land sector opportunities through biodiversity offsets, land stewardship benefits, market-chain accreditation schemes fostering greater sustainability in the pastoral sector and associated environmental services enterprises. Such a vision, however, contrasts starkly with the Australian Governments’ blinkered view of what the northern development agenda should look like – a business-as-usual focus on big infrastructure developments including dams, and other long-cherished but unfounded development myths such as food bowls and extensive irrigable agriculture. By contrast, we need to foster a development agenda that is cognisant of and responsive to regional, especially indigenous cultural and demographic realities; as well as significant commercial and environmental challenges facing the major extensive land-use sector – the northern pastoral industry. The development of a diversified, sustainable land sector economy that nurtures local communities and complements pastoral enterprises on the great majority of country (with, at best, marginal grazing enterprise potential) has obvious long-term soil, water, biodiversity – and social and cultural benefits. For the large-sized pastoral properties that characterise the northern landscape, why not encourage pastoral activities on those limited portions best suited to extensive (as opposed to intensive, irrigated or grain-fed) grazing systems (e.g. fertile grasslands), while incentivising the provision of ecosystem services (e.g. carbon farming; soil, water and biodiversity conservation) on the relatively ‘unproductive’, fire-prone and generally biodiverse very extensive remainder? In this context, public expenditures on Indigenous Ranger Programs, Indigenous Protected Areas, and the like, might better consider how those investments could also be used to kick-start locally owned and managed land sector enterprises. Sounds unrealistic and improbable? That is what we were told when we started exploring market-based savanna burning opportunities less than 20 years ago. Jeremy Russell-Smith is Adjunct Professor with the Charles Darwin University (Darwin, NT 0909, Australia; Tel: +61 4 4720 0927).