The Grès supérieurs is not dramatic rock. Sandstone, siltstone and mudstone, stacked across the Savannakhet Basin and stained a dull iron red, it looks like ground you would drive past without a second glance. The story is in the grain. These are river deposits, laid down between roughly 120 and 100 million years ago in the Aptian and Albian stages of the Early Cretaceous, when this corner of Asia was a broad lowland threaded with channels. Sand buried whatever the water carried, and kept it. Across the border in Thailand the same interval survives as the Khok Kruat Formation, close enough in age and character that geologists treat the two as equivalents. What is preserved here, in other words, is a floodplain — compressed, tilted, and waiting.

Hoffet worked quickly and published what he had. In 1942 he described titanosaurian bones from what he called Bas-Laos, lower Laos, and over his career he attached names to the animals he pulled from the sandstone: Titanosaurus falloti for a sauropod, Mandschurosaurus laosensis for a large plant-eater. Neither name survived the century intact. Titanosaurus has since been dismantled by paleontologists as a catch-all genus, and falloti is now carried as an indeterminate sauropod. Mandschurosaurus laosensis has been reduced to an unidentified iguanodontian. None of this is a failure of Hoffet's work. He was naming animals from fragments, in a country with no comparative collections to check them against, decades before anyone knew what else had lived here.

In 1999 a team led by the French paleontologist Ronan Allain published a new sauropod from the formation and gave it a name that settles an old debt: Tangvayosaurus hoffeti. The genus honours Tang Vay, the village in Savannakhet Province near where the bones emerged; the species honours the geologist who got there first. At around fifteen metres long, it was a serious animal — a long-necked browser working its way through those vanished river valleys. More importantly, it gave the Grès supérieurs something it had never had: a well-founded, diagnosable dinosaur rather than a list of contested names.

Then came the animal that made the formation internationally famous. In 2012 Allain, working with T. Xaisanavong, P. Richir and B. Khentavong, described Ichthyovenator laosensis from a partial skeleton recovered at Ban Kalum — vertebrae, hips and ribs lifted from a patch of red sandstone smaller than two square metres. It was a spinosaurid, one of the long-snouted, crocodile-jawed theropods that hunted fish. Isolated spinosaurid teeth had turned up in Asia before, but this was the first definitive Asian member of the family known from real skeletal material. It was also strange by the standards of its own group. Where its African relatives carried a single tall sail, Ichthyovenator's dipped over the hips and rose again in a sinusoidal double curve, the twelfth back spine standing 54.6 centimetres high. Estimates put the living animal somewhere between eight and ten metres long, and around two tonnes.

Dinosaurs take the headlines, but the formation preserves a fuller world than that. A gar-like fish, Lanxangichthys alticephalus, swam these channels — a basal lepisosteiform described from Cretaceous rocks in both Laos and Thailand. A freshwater turtle, Shachemys laosiana, has come out of the same beds. Trackways record sauropods and small ornithopods crossing wet sediment, footprints being the rare fossil that catches an animal alive rather than dead. Most of the unit remains unworked. Excavation at Ban Kalum has continued to produce more of Ichthyovenator since the original description, which is a fair summary of the whole formation: what has been found is a fraction of what is down there, and the digging is patient work.