[3] The Antarctic toothfish was first formally described in 1937 from the English ichthyologist John Roxborough Norman with the type locality given as off MacRobertson Land at 66°45'S, 62°03'E in Antarctica.

This loosely schooling species is also a major prey of Adélie (Pygoscelis adeliae) and emperor penguins (Aptenodytes forsteri), Weddell seals (Leptonychotes weddellii) and Antarctic minke whales (Balaenoptera bonaerensis).

Toothfish that are dwelling on the bottom, particularly those caught during the summer on the continental slope, eat mainly grenadiers (Macrouridae), but also feed on other smaller fish species and skates (Raja spp.).

[18] Aging data indicate Antarctic toothfish are relatively fast-growing when young, but then growth slows later in life.

[23][24] Large, mature, older fish have been caught among the seamounts of the Pacific-Antarctic Ridge, a location thus thought to be important for spawning.

The Antarctic toothfish has a lightweight, partially cartilaginous skeleton, lacks a swim bladder, and has fatty deposits which act as a stored energy source, particularly during spawning.

Antarctic toothfish have vision and lateral line systems well adapted to find prey in low light levels.

Antarctic toothfish also have a very well developed sense of smell,[27] which is why they are easily caught by baited hooks and also scavenge the remains of penguins killed by other predators.

It is noteworthy, like most other Antarctic notothenioids, for producing antifreeze glycoproteins, a feature not seen in its closest relative, the Patagonian toothfish, which typically inhabits slightly warmer waters.

The presence of antifreeze glycoproteins allows the Antarctic toothfish (and other notothenioids) to thrive in subzero waters of the Southern Ocean surrounding Antarctica.

The existence of this fishery in the Ross Sea, the area where most Antarctic toothfish are caught, is very contentious - the main argument proposed for this is the lack of accurate population parameters, such as original stock size, fecundity, and recruitment.

Moreover, the main fishing grounds are presumed by some researchers to cover the area through which the entire stock of Antarctic toothfish pass.

[31] Current spawning stock biomass for Antarctic toothfish in the Ross Sea Region is estimated to be at 75% of the pre-exploitation level (95% Bayesian probability interval 71–78%), well above the 50% target reference point.

[32][34] More recent studies have confirmed visual sightings of Weddell seals and Type-C killer whales holding and consuming large toothfish in the McMurdo Sound area and raise questions over the previously assumed importance of assumed dominance of Antarctic silverfish (Pleuragramma antarcticum) in the diet of Weddell seal and Type-C killer whales.

[35][36] These reports highlight the importance of managing this fishery in the best interests of the ecosystem by continuing to collect information on both Antarctic toothfish life history and the interaction of that species with predators and prey.

An important[citation needed] research programme in this regard is the annual 'Shelf' survey carried out annually since 2012, which is designed to monitor the abundance of subadult Antarctic toothfish in areas where subadult-sized fish have been regularly found (e.g., in the southern Ross Sea) has been designed provide data to better estimate recruitment variability and provide an important early-warning signal of changes in toothfish recruitment.

[37] Research has provided evidence for long-distance migrations of type-C killer whales between the Ross Sea and New Zealand waters, indicating a much wider range that had been postulated by a number of scientists.

One adult female type-C killer whale has been seen in both New Zealand waters and McMurdo Sound, Antarctica, and a high large proportion of type-C killer whales sighted in McMurdo Sound have scars caused by cookiecutter sharks that are currently assumed to be limited to north of 50°S.

A common misunderstanding[citation needed] of the CCAMLR decision rules is an assumption that the decline in population size will follow a clear trajectory from the starting year to a point 35 years later when the stock size will reach 50% of pre-exploitation levels and an assumption that no feedback occurs during each assessment.