The Most Accurate T. rex Ever Painted Was Chubby, Bare and Built to Be Argued With

The T. rex that gets called the most accurate ever painted is thick through the middle, smooth-skinned and completely bare. No fuzz. Small arms held close to the body. A jaw built so the teeth disappear when the mouth closes.

RJ Palmer finished it in October 2018 after 320 hours of work, commissioned by the developers of Saurian, a survival game set in the Hell Creek ecosystem of 66 million years ago. The image went on show in Picturing the Past, a juried paleoart exhibition at the New Mexico Museum of Natural History and Science in Albuquerque that ran from October 2018 into January 2019.

It's worth looking at again, not because the painting is correct, but because of how it was built and how quickly parts of it were overtaken. That's the useful story here. A reconstruction of an extinct animal is an argument dressed as a portrait, and the good ones show you what they're arguing from.

Built from the skeleton up

Palmer read more than 20 published papers on the animal, a dozen of them on musculature alone. 'I left no stone unturned,' he told the San Francisco Chronicle. 'The process was extremely grueling. I spent months trying to figure out the shapes of muscles from paragraphs of text filled with cumbersome anatomical terminology as my only guide,' he said later.

Scott Hartman, a palaeobiologist known for skeletal reconstructions, consulted on the anatomy. 'The team worked from the skeleton up, spent months getting the muscles right, and the results are now available for everyone to see,' he said. 'The amount of time invested and attention to detail blew me away.'

What that produced was a list of small decisions, each traceable to something. Keratin plates along the back of the neck, following a pattern seen in birds. Extra tissue along the jawline to hide the teeth when the mouth closed, inferred from small holes in the bone that would have fed nutrients to soft tissue. Blunt claws on the feet, worn down by contact with the ground, and sharp ones on the hands, which barely touched anything.

None of it was peer-reviewed, and nobody involved pretended otherwise. 'As with any reconstruction there are of course some details we can't know, and competing interpretations that are equally likely,' Hartman said. 'But I can honestly say this was the most exhaustive attempt at restoring an extinct animal that I've worked on.'

Palmer's own framing was simpler. 'I wanted the world to see T rex as an animal, not as some movie monster,' he said.

The feather fight was already running

Five months before the painting appeared, a team led by Phil Bell, then at the University of New England in Australia, published fossil skin from Tyrannosaurus and four close relatives in Biology Letters. The T. rex material came from a Montana specimen nicknamed Wyrex, plus samples held at the Houston Museum of Natural Science: patches of hide from the neck, pelvis and tail, alongside skin from Albertosaurus, Daspletosaurus, Gorgosaurus and Tarbosaurus.

All of it was scaly. Small, pebbly, reptile-grade scales, with no filament or feather impressions anywhere in the sample. The paper went further than the skin itself. Body size in tyrannosauroids shows two separate jumps to giant size, in Yutyrannus and in the tyrannosaurids, and the authors argued that the dense feathering seen in earlier tyrannosauroids had been lost by the Albian, before the family's largest members appeared. If any plumage survived on T. rex, they wrote, it was probably confined to the back.

The BBC's coverage carried the immediate objection from Steve Brusatte at the University of Edinburgh. 'It takes inconceivable good luck to preserve feathers in fossils,' he said. 'Just because we don't see them doesn't mean they weren't there.' Matthew Baron, then at Cambridge, offered the ostrich: feathers across most of the body, scales on the legs. Bell's own answer to that was careful. 'It's a reasonable theory, but there's more that we don't know than what we do.'

The full study is short and readable, and worth the twenty minutes if you want to see what fossil integument actually looks like under a lens. What it doesn't do is settle the argument. Palmer's naked T. rex took a side, and it took the side with the physical evidence then available, which is the defensible way to make that call. It still left the animal's back open to dispute.

Photo by El Swaggy on Unsplash

Five years later, the face changed again

In March 2023, Thomas Cullen at Auburn University and colleagues published a study in Science that went straight at the mouth. The question was whether theropod teeth would even fit behind lips. They sectioned tooth enamel, measured how thickness differed between the inner and outer faces of teeth in living crocodilians and monitor lizards, and compared tooth size against skull length across a wide reptile sample.

Crocodiles have no lips and live in water, and their outer enamel wears faster. The dinosaur teeth didn't show that pattern. Enamel was evenly thick on both sides. And the teeth of Sue, the Field Museum's T. rex, were no longer relative to skull size than a monitor lizard's, so a lip covering them was plausible on geometric grounds rather than aesthetic ones. The published conclusion was a scaly, lizard-like lip line, closer to a Komodo dragon than to the crocodile smile that had been standard for a century.

Ali Nabavizadeh of the University of Pennsylvania, who wasn't involved in the work, called it an excellent case for soft tissue covering the mouth in non-avian theropods. Cullen's own line, reported by Nature, was blunter: lips, or feathers, have no bearing on whether an animal is frightening.

Palmer's 2018 painting had already put tissue around the jaw for exactly that reason, working from foramina in the bone rather than from enamel histology. The image was right about the direction five years before the data arrived. That's not luck. It's what happens when an artist builds from anatomy instead of from other artists.

The bulk was a guess too, and still is

The same problem applies to the body. In July 2024, Jordan Mallon at the Canadian Museum of Nature and David Hone at Queen Mary University of London modelled an entire T. rex population along an updated growth curve — 140 million simulated individuals — to ask how likely it is that the fossils in museum drawers include the biggest ones.

The answer is: not likely at all. The largest known specimens sit around the 99th percentile for body mass, and finding an individual in the top 99.99% would take another thousand years of excavation at the current rate. The biggest animal that could plausibly have existed, a one-in-2.5-billion individual by their published model, comes out around 15 tonnes against Scotty's 8.8, and 15 metres long against 12.

Thomas Carr, who directs the Institute of Paleontology at Carthage College, told CBC the estimate gives other researchers a benchmark for future finds. 'What we think is the maximum size for any dinosaur species is almost certainly wrong simply by virtue of the incompleteness of the fossil record,' he said.

So the chubby silhouette is a reasonable mid-range animal, drawn well and drawn from bone. It isn't a portrait of the largest one that ever lived, or the leanest, or the youngest. Nobody claimed it was. The label did that work on Palmer's behalf.

Where the phrase 'most accurate' comes from

The claim has a history that predates Palmer's painting. In 2012, John Conway, C. M. Kosemen and Darren Naish published All Yesterdays, a book aimed at what they called shrink-wrapping: skin pulled so tight over a skeleton that individual bones show through the living animal. Their argument was simple and hard to answer. No living vertebrate looks like that. Fat, loose skin, fluff and colour cover the bones, and reconstructions that leave those out are making a stylistic choice, not a scientific one.

The skeletal reconstructions in that book were drawn by Scott Hartman, the same man who spent months on the muscles for Palmer's T. rex six years later. Naish later wrote about how much of any reconstruction is judgement rather than measurement, and about how the field's conventions hardened into habits nobody had tested, in a piece for Scientific American that reads now as a warning about exactly this kind of headline.

Criticism of the label hasn't faded. In June 2025 the palaeoartist Matt Dempsey argued that the Jurassic Park T. rex works better as a caricature of the animal than much current palaeoart does as a likeness, because popular reconstructions smooth away the bony landmarks that make a skull recognisable. It's a fair hit. Soft-tissue accuracy is easy to claim and hard for a reader to check. A skull with the right ridges and openings on it can be checked by anyone with a photograph of the fossil.

Mark Witton, another consultant on the project, put the challenge to critics directly: 'can you do something better with the data we have to hand?' A decade on, that's still the right question, and the answer from most people who complain about reconstructions is no.

Photo by Olga Deeva on Unsplash

What to check when a museum shows you a dinosaur

Three questions do most of the work. What is the claim built on — fossil skin, bone, teeth, trackways, or inference from living relatives? What did the artist have to invent, and is the invention flagged anywhere a visitor can see it? And has any of it been tested since?

Palmer's T. rex holds up well on the first two. The skin comes from skin impressions. The jaw tissue came from holes in bone. The claws follow from how the feet met the ground. The feathers are an absence, and absences are the weakest evidence in the file — a point Brusatte made in 2017 that nobody has since overturned.

Expect the next version to move again. It usually moves because someone sectioned a tooth, measured a femur or found a new patch of fossil hide, not because a painter changed their mind about what looks cool.