Labyrinthodonts were giant amphibians distantly related to modern newts and salamanders. In outward appearance labyrinthodonts looked much like modern salamanders, but with two main differences: they had jaws full of sharp teeth; and they often grew to enormous sizes. Labyrinthodonts are named for the pattern of infolding of the dentine and enamel of the teeth, which resemble a maze (or labyrinth).

Most species became extinct during the Permian and Triassic periods, even before dinosaurs or mammals evolved in some cases. However, whereas most labyrinthodonts died off around 250 million years ago at the latest, the Australian animals lived on until the Early Cretaceous, an extra 130 million years. Australia is the best place to find fossils of labyrinthodonts. In some rocks called the Arcadia Formation in Queensland there are the fossils of more sorts of labyrinthodonts than anywhere else in the world.

The smaller labyrinthodonts ate insects, but the bigger ones ate larger animals. The many sharp teeth they had were useful for catching such things as fish and perhaps unwary small dinosaurs that got too close to the water’s edge. But when the teeth of one good-sized labyrinthodont called Siderops kehli were carefully examined, all that was found were the fossilised remains of millipedes and a piece of the backbone of another labyrinthodont.

However, the really big creature that lived alongside the polar dinosaurs, Koolasuchus cleelandi, must have eaten something more substantial than this to sustain itself, and probably lived on slow-moving lung fish with the occasional young dinosaur for a change.

We can be pretty confident that they laid eggs, but we don’t know if they had a tadpole stage of life like most other amphibians, which have all sorts of odd ways of breeding. So it's difficult to know for sure how labyrinthodonts bred and whether they did have tadpoles or larvae, but we do know that reproducing this way happened with other amphibians of the time. In China there is a fossil clearly showing the larva of a salamander with gills and no legs, but with a fin-like tail for swimming.

Some temnospondyls were completely aquatic, with external gills much like an axolotl, or "Mexican walking fish". Some others became as terrestrial as reptiles, only returning to the water to breed. Temnospondyls had large heads with advanced hearing abilities similar to frogs, capable of hearing sounds moving through air. Channels in the bone of the skull roof helped them to detect vibrations within the water as well.

All of the Australian animals belong to the Brachyopoidea; indeed it seems that when Australia was part of Pangea the brachyopoids originated there, eventually spreading to other parts of the ancient supercontinent. Australia would have been the ancestral home of Brachyopoid labyrinthodonts. The Australian examples all seem to belong to the Temnospondyls, one of the most diverse and successful orders of labyrinthodonts, which included the brachyopoids.

There are many species of labyrinthodont known from Australia, but only three that post-date the Triassic Period. Austropelor wadleyi is known only from a fragment of lower jaw, from Jurassic deposits in Queensland. When first found, and published as a Jurassic labyrinthodont, there was such an outcry by other researchers that the deposits were re-dated as Triassic based on the presence of this one bone. It was not until the near-complete Siderops material was found, and after the Austropelor deposits were dated to the Jurassic via pollen analysis, that people began to believe that some labyrinthodonts had survived the Triassic period.

Siderops kehli lived in the Early Jurassic of south-central Queensland. It is known from an almost complete skeleton preserved in iron stone (hence the name Siderops: "iron face"). In fact the surrounding matrix was harder than the bones themselves, and the skeleton took quite a long time to be prepared. It measured about 2.5 metres long, with a shovel-shaped head 57cm long and 67cm wide. At the time of discovery it was one of the largest known labyrinthodont species. Its teeth were truly impressive, especially by labyrinthodont standards.

On the palate a row of about 100 teeth were small towards the back, but rapidly increased in size to 30mm towards the front. Inside these was a continuous row of approximately 80 teeth, a bit bigger than the outer row and projecting beneath them. In this inner row of teeth were three sets of tusks on either side, the largest of which extends almost 70mm below the jaw line. The teeth were shaped like those of other aggressive carnivore species, such as monitor lizards. On each side of the tooth a sharp ridge extended towards the tooth tip. When the upper and lower jaws were brought together, the teeth slid past each other like a series of scissor-like blades. The surface of the palate in the area of the palatal bones was covered by tiny 1-2mm denticles, similar to shark skin. Once Siderops got something into its mouth, there was little hope of escape.

With its large head and impressive teeth, but with a relatively weak body, it probably hunted food underwater, and perhaps took land animals at the water's edge, much like a crocodile would. To capture small prey in the water, all it needed to do was to open its huge mouth, and the in-rush of water would have sucked prey straight in. That it had teeth at all tends to indicate that Siderops was not just content on sucking in small prey.

The third species, Koolasuchus cleelandi, was the latest surviving labyrinthodont known in Australia, dating to about 120 million years ago. It was named in honour of Leslie Kool, a preparator whose job it is to remove the fossils found in the Victorian sites from the rock that surrounds them. The specific name honours Mike Cleeland, who found the left and right mandibles of the type specimen. The name is also ironic, since it lived in a "cool" climate, and probably filled the same niche as the absent crocodiles (suchus = crocodile). Known mostly from jaw fragments up to 80cm long found at San Remo, its total length has been estimated at between four and five metres. There seemed to be a general trend towards increasing size amongst labyrinthodonts, with the latest surviving species being the largest of all.

It is thought that in other parts of the world competition from crocodiles wiped out most labyrinthodont species. However during the Early Cretaceous southern Victoria would have been within the Antarctic circle, too cold for crocodiles. The climate seems to have warmed up a bit by 110 MYA, though. The remains of crocodiles have been found in these deposits—but no labyrinthodonts. Koolasuchus may well have been the last of its kind anywhere in the world.