The taxonomy of the Isoodon obesulus / auratus group has been confused. Pope et al. (2001) and Zenger et al. (2005) suggested that I. auratus is possibly conspecific with I. obesulus, but the two forms were allopatric even if the late Pleistocene subfossil material is included (McKenzie et al. 2008), and Westerman et al. (2012) showed that the two species were distinct. The status of subspecies of I. auratus has also been unclear, but Westerman et al. (2012) agreed that I. a. arnhemensis should be merged with I. a. auratus and that I. a. barrowensis was also close to I. a. auratus. Recent genetic analysis recorded statistically significant though unequally distributed levels of genetic diversity among the three recognised subspecies (Rick et al. 2023), but only two subspecies, I. a. barrowensis and I. a. auratus (merged with I. a. arnhemensis) are listed independently as Vulnerable under the Environment Protection and Biodiversity Conservation (EPBC) Act 1999. Travouillon and Phillips (2018) comprehensively reviewed the taxonomy of bandicoots and bilbies and concluded that I. auratus was a full species, related to I. fusciventer and I. obesulus.

We have evaluated I. auratus as a single species with no subspecies and have evaluated I. peninsulae and I. fusciventer as separate species.

The species is listed Least Concern. Conservation actions over the last 10-20 years have led to some improvements in the status of this species. Although population trends are not well resolved, it is likely that any decline in the residual subpopulation of mainland Kimberley has been moderated by improved fire management. Subpopulation size is probably stable (but varying with rainfall) on the 7 islands in which original subpopulations persisted, and reintroductions have led to increasing subpopulations on 4 islands and three mainland sites – although two of these sites within ‘mainland islands’ (predator exclosures) have been established for less than 5 years, so are not included in this status assessment. Given the likely subpopulation trends across sites, it is implausible that criterion A (more than 30% decline over 10 years) is met. Translocations over the last 10-20 years have increased the number of locations to more than 10. The species does not undergo extreme fluctuations and is not severely fragmented. The total population size is not well resolved, but substantially exceeds 10,000 individuals, so criterion C is not met. Near Threatened is may be more appropriate than Least Concern as the number of locations may reduce to 10 or fewer if some of the translocated subpopulations do not persist; and area of occupancy (AOO) is less than 2,000 km² and there may be a continuing decline in at least habitat quality due to fire (although fire regimes are now better managed) and number of mature individuals due to predation by feral Cats on the Kimberley mainland.

The Golden Bandicoot is endemic to Australia. The species was formerly very widespread in western (Friend 1990), central (Burbidge et al. 1988) and northern Australia extending to western Queensland, New South Wales (Ellis et al. 1991) and Victoria. It is now extinct on the mainland except in a few locations in the north-west Kimberley between Mitchell Plateau in the north (McKenzie et al. 2023) - although Wunambal Gaambera Aboriginal Corporation (WGAC) have camera trap records from sandstone areas around Rocky Cove that are likely to be Golden Bandicoot but for which identification is not conclusive based on camera images alone - and Artesian Range in the south (Rick et al. 2023). It occurs on Barrow (235 km²) (including the small Boomerang Island, connected to Barrow at low tide) and Middle (3.5 km²) Islands in the Pilbara, Lachlan (12 km²), Augustus (190 km²), Storr (19 km²) and Uwins (32.5 km²) Islands in the Kimberley (Gibson and McKenzie 2012), and Marchinbar Island (210 km²) in the Northern Territory (Woinarski et al. 1999)

It was extinct on Hermite Island, Western Australia (WA) (10 km²; Burbidge 1971), but was reintroduced there from Barrow Island (WA) in 2011 (Thomas et al. 2014). Golden Bandicoots were introduced to Doole Island, WA in 2010 (Ottewell et al. 2014, Rayner et al. 2021), and to Guluwuru (41 km²: 2007) and Raragala (94 km²; 2008-09) Islands in the Wessel group, Northern Territory (Palmer 2009).

Golden Bandicoots were reintroduced to a fenced reserve free of exotic predators (‘mainland island’) on the Matuwa Kurrara Kurrara National Park WA (11 km²) in 2010 (Lohr et al. 2021). Escapees from the enclosure are occasionally observed several kilometres from the fenced area. In 2022, Golden Bandicoots from the Matuwa Kurrara Kurrara population were reintroduced to the Wild Deserts fenced reserve (20 km²) in Sturt National Park, western New South Wales (Moro et al. 2024). It is planned in 2025 that bandicoots will be released beyond the fence into a 100 km² semi-bounded area at Wild Deserts. In 2023, the Australian Wildlife Conservancy (AWC) reintroduced Golden Bandicoots from the AWC Charnley River – Artesian Range Wildlife Sanctuary and Barrow Island to an exclosure in the AWC Newhaven Sanctuary (94.5 km²) in the Northern Territory. Additional translocations are proposed e.g., Moro et al. (2024).

An earlier reintroduction to part of the Gibson Desert, WA, failed due to predation by feral Cats (Christensen and Burrows 1994).

There is no robust estimate of population size. It is abundant on Barrow Island (densities of up to 10 adults ha^-1 in sandy areas, and population size more than 20,000 individuals: McKenzie et al. 2023) and relatively abundant on some other islands, but is rare and probably declining in mainland Kimberley. The subpopulation on Marchinbar Island was estimated at 1400 individuals in 1994-95 (Woinarski et al. 2007), based on limited density information derived from radio-tracking and intensive trapping (Southgate et al. 1996). It is relatively abundant on some Kimberley islands.

Subpopulations have been reintroduced to fenced mainland islands (exclosures) free of introduced predators in central WA (Matuwa Kurrara Kurrara National Park), western New South Wales (Sturt National Park in 2022), and central Northern Territory (Newhaven). The subpopulation in Western Australia having been established for 14 years. Golden Bandicoots have been observed outside the fenced reserve in Western Australia on multiple occasions, but if there is a persisting subpopulation outside the reserve it remains at extremely low density (Lohr et al. 2021). The abundance of Golden Bandicoot inside the exclosure cannot be calculated accurately due to trap interference by Boodies (Bettongia lesueur).

Population trends over the last 10-20 years are not well resolved. Subpopulations on Kimberley islands (which are largely devoid of threats) are unknown but likely to have been stable. Declines, if any, on the Kimberley mainland may have been counterbalanced by the establishment of new subpopulations on Doole, Raragala and Guluwuru Islands, and on mainland sites at Matuwa Kurrara Kurrara, Sturt and Newhaven, although reintroductions to the latter two sites are too recent to presume that these subpopulations are securely established. There has been little monitoring of the subpopulations on the Wessel islands (Raragala and Guluwuru) or Kimberley islands.

The Golden Bandicoot is nocturnal. On Barrow Island, Golden Bandicoots shelter in medium-sized or large spinifex Triodia spp. hummocks and in limestone caves and crevices. In Matuwa Kurrara Kurrara National Park they may shelter inside Boodie (Bettongia lesueur) warrens or dig their own nests under leaf litter gathered at the centre of mallee-form Eucalyptus sp. or under the shrubs of Acacia sp. (Cheryl Lohr pers. comm.  2024). In Sturt National Park they make nests under native grass tussocks and species such as Lechenaultia divaricata, as well as using Bilby Macrotis lagotis burrows.

Food is obtained by digging and foraging and includes insects, small reptiles and small mammals, sea turtle eggs, carrion, roots and tubers (Radford 2012). Females give birth throughout the year to litters of two or three young, all of which may survive, but the proportion of females with young increases after heavy rainfall. In the Kimberley, diet includes insects and plant material.

On Marchinbar Island it occurs mainly in heath and shrublands on sandstone and individuals maintain overlapping home ranges of 12-35 ha and their diet includes a broad range of invertebrates (Southgate et al. 1996). Radio-tracked individuals moved regularly between shelter sites over a 600 m long riparian strip grading to grassy foothills. At some sites the Golden Bandicoot was sympatric with the Northern Brown Bandicoot I. macrourus.

Observations of Isoodon auratus in the north-west Kimberley, Western Australia, suggest that it is often found adjacent to dissected sandstone, usually on flats adjacent to sandstone associated with Triodia spp. grass cover, rather than in the really rugged rocky sandstone habitats and, at Mt Trafalgar in Prince Regent National Park, it occurs adjacent to sandstone scree and rainforest at the base of the scarp in volcanic woodland habitat Eucalyptus tectifica/tussock grass vegetation. So, the species is not restricted to sandstone. (I. Radford pers. comm. 2024). At Mitchell Plateau, Golden Bandicoot populations occur side by side within 100 m of Northern Brown Bandicoot Isoodon macrourus populations at Surveyors Pool monitoring sites, but no I. auratus were captured at the Northern Brown Bandicoot site and no Northern Brown Bandicoots were captured at the Golden Bandicoot site. This may be ecological partitioning, with Northern Brown Bandicoots preferring the riparian vegetation (tall dense riparian grass) and goldens the sandy spinifex sites adjacent to nearby sandstone habitats. However, at Mt Trafalgar in volcanic woodlands, and at Bachsten Creek in sandstone woodlands, both bandicoot species are captured in the same habitats (I. Radford pers. comm. 2024). In some circumstances Golden Bandicoots can benefit from small-scale burns. Isoodon auratus increased its consumption of vertebrate prey after fire at the Mitchell Plateau (Radford 2012).

At Matuwa, Golden Bandicoots were caught in all available habitats inside the fenced exclosure but preferred mixed shrubland with medium density Triodia spp. hummocks (Lohr et al. 2021). At Sturt National Park, bandicoots prefer the dune habitats but are also known to use the swale habitats (unpublished radiotracking studies).

Two species of nematode have been identified from I. auratus, Linstowinema warringtoni and Labiobulura peramelis, but are not known to threaten the species (Smales et al. 2024).

On mainland Kimberley, Golden Bandicoots are threatened by Feral Cats Felis catus combined with changing fire regimes. These two threats work synergistically to the detriment of the species, whereby ground cover reduced by fire makes bandicoots more susceptible to predation (Palmer et al. 2003). Feral dogs on Marchinbar may have been a minor threat, but were eradicated in 2004-2005 (Palmer and Woinarski 2006). There is a minor risk of introduced predators establishing on Barrow and Middle Islands; however, after approval was given to construct a major LNG plant on Barrow Island, a comprehensive quarantine management system was developed, approved and implemented (Scott et al. 2017). Lower densities on Middle Island (relative to Barrow Island) may have been due to competition from the introduced Black Rat (Rattus rattus), however, these were eradicated in 1993 and since then numbers appear to have increased (Morris 2002, McKenzie et al. 2023).

The reasons for past declines of the Golden Bandicoot are predation by feral Cats and Red Foxes Vulpes vulpes, and inappropriate fire regimes (the Aboriginal fire regime of mostly small fires and a heterogeneous landscape has been replaced by very large, hot wild fires). Improving fire management, often by Aboriginal rangers, is being implemented in some areas. Feral Cat control likely contributes to the occasional observation of Golden Bandicoots outside the fenced reserve on Matuwa Kurrara Kurrara National Park (Lohr et al. 2020).

A series of translocations over the last 10-20 years have established populations on three islands without invasive predators, and to three mainland sites at which invasive predators are excluded by fencing. These actions have led to increases in the number of locations and subpopulations. The residual subpopulation on the Kimberley mainland is being managed through imposition of a fine-scale fire regime by Indigenous Ranger groups, the Australian Wildlife Conservancy and DBCA. Some of these actions were undertaken under the guidance of a multi-species (with Golden-backed Tree-rat Mesembriomys macrurus) recovery plan (Palmer et al. 2003). Given the recent series of successful translocations, the long-term trend for decline in this species has probably been staunched and the species is on a pathway to recovery (Woinarski et al. 2023a). Nonetheless, this recovery still falls far short of return to the distributional extent and population size it had prior to European colonisation of Australia, and will be partly dependent upon ongoing investment and maintenance of predator-exclusion fencing at mainland sites and biosecurity for island sites (Woinarski et al. 2023b, Read et al. 2023), plus ongoing work to establish translocated Golden Bandicoot subpopulations beyond fences using a combination of improved predator control and in situ predator training (Blumstein et al. 2019).

Barrow and Middle Islands are nature reserves vested in the Parks and Wildlife Commission of Western Australia and managed by the Department of Biodiversity Conservation and Attractions. Chevron Australia is responsible for management of the oilfield and liquefied natural gas plant and their impacts under environmental approval from the State and Commonwealth governments. Most mainland places and Kimberley islands on which Golden Bandicoots occur are Aboriginal land. Reintroduction to Hermite Island (part of the Montebello Islands Conservation Park) from Barrow Island, and introductions from Marchinbar to Guluwuru and Raragala Islands (all now within an Indigenous Protected Area) have occurred and are being monitored (albeit sporadically). Subpopulations in fenced reserves are intensively monitored. Matuwa Kurrara Kurrara National Park was created under an Indigenous Land Use Agreement between Tarlka Matuwa Piarku Aboriginal Corporation and the Department of Biodiversity Conservation and Attractions. A major quarantine program to minimise risks of establishment of new threats on Barrow Island is in place (Scott et al. 2017) .

Isoodon auratus is one of the larger marsupial species that benefitted from introduction of burning mosaics, as measured by increases in trap success (Radford et al. 2020).

 The Golden Bandicoot is listed as Vulnerable (for each of the two subspecies) under Australian environmental legislation.