Did Yutyrannus REALLY Have Feathers?

Yutyrannus was discovered in China in 2012 and made headlines as the first large dinosaur found with fossil filaments — evidence that’s since made feather-covered tyrannosaurs a fixture of documentaries, books, toys, and paleoart. But the two known specimens preserve filaments only in patches near the neck, back, arm, foot, and tail, and the preservation is poor enough that none of the usual tests (molecular studies, cross-sections, imaging under different light) were ever run on them. This video traces how these filaments — and similar structures found on Sinosauropteryx, Psittacosaurus, and Beipiaosaurus — came to be labeled “protofeathers,” and weighs that history against three more recent, much better-preserved discoveries (Kulindadromeus, Juravenator, and Haolong dongi) that increasingly point toward alternatives like collagen fibers, keratinous spines, or other non-branching integument rather than true feathers.

Overlooked Anatomy: Arms, Skull, and Three Crests

Along the way, Joel Leineweber paints a new reconstruction of Yutyrannus that leans on lesser-known details of the fossil. Unlike most tyrannosaurs, Yutyrannus had larger arms with three fingers, and its skull looks different enough that some researchers think it may be more closely related to carcharodontosaurs than to typical tyrannosaurs. Most paleoart is so focused on showing off the feathers that it misses another distinctive feature: Yutyrannus had three head crests, a prominent midline crest plus one over each eye, which would have made for a striking display. Since the fossil’s keratin sheaths weren’t preserved, there’s no way to know how far those crests actually extended — which is what inspired this reconstruction’s deliberately dragon-like pose.

How Filaments Became “Feathers”

To understand why Yutyrannus’s filaments get called feathers at all, the video walks back through the history of the claim. In 1860, a single feather belonging to Archaeopteryx was discovered in Germany — a completely modern-looking feather, with a quill and organized barbs, found in Jurassic rock. That discovery kicked off a search for how feathers evolved: Archaeopteryx itself turned up in late Jurassic layers, and Anchiornis, with modern-looking wing feathers, was later found in even older middle Jurassic rock.

Then, in 1996, researchers described Sinosauropteryx, a small compsognathid with fuzzy filament structures down its neck, back, and tail. Those filaments were immediately labeled “protofeathers.” John Ostrom, who had spent his career trying to prove birds evolved from dinosaurs, reportedly began to cry and nearly fell to the floor when he saw the fuzzy Sinosauropteryx fossil — a reaction that says more about how invested researchers were in the idea than about the actual evidence.

In 2002, Psittacosaurus turned up with long spines at the base of its tail — the rest of its body was covered in ordinary reptilian scales, with just that one spot bearing the spines. These had no barbs and no branching structure, so researchers called them “bristles” instead of feathers, saying only that they might be related to filaments. One detail stands out: these spines appear to attach directly to the tail vertebrae rather than sitting embedded in skin like scales. A few flight feathers in living birds do attach to bone, but what’s preserved on Psittacosaurus still has no branching structure like a real feather.

In 2009, Beipiaosaurus — a therizinosaur — was described with filaments and spiny structures on its neck and tail. Researchers labeled two types of “feathers” on this animal even though neither branched or had barbs: ordinary filaments like those on Sinosauropteryx, and what they called “elongated broad filamentous feathers,” which look a lot like spines. Beipiaosaurus was named from two disconnected fossil slabs, though, and most other therizinosaurs are known from so few articulated specimens that scientists can’t be fully certain the bones all belong to the same species. A more complete, fully articulated therizinosaur, Jianchangosaurus, was described in 2013 with faint filament impressions near the base of its neck.

Why Scientists Called Them Feathers Anyway

By this point, researchers were using the word “feather” for all of these structures, and the video reviews the reasons they gave: melanosomes (pigment structures that look identical to bacteria, and that also occur in ordinary skin and scales — a resemblance close enough that researcher Mary Schweitzer reportedly worried “she’s going to ruin us”), hollow structures (which are only preserved two-dimensionally, so their original shape can’t really be confirmed), and beta-keratin (a protein that can be denatured between beta and alpha forms under heat and pressure). Despite these arguments, scientists like Dr. Alan Feduccia have concluded that most of these filaments are more likely decaying collagen fibers or other internal structures — a conclusion the video suggests gets downplayed because researchers pushing the dinosaur-to-bird narrative need protofeathers to exist. For Yutyrannus specifically, preservation was so poor that none of this supposed evidence was even found, yet the word “feather” still appears 26 times in the paper describing it. Some of the Yutyrannus structures are also longer than collagen fibers typically get, which is one reason this reconstruction interprets them as spines more similar to those on Psittacosaurus.

Three New Discoveries That Changed the Picture

Three more recent, much better-preserved fossils bring new evidence to this discussion. Kulindadromeus, discovered in 2014 in middle Jurassic rock in Russia, had structures so unusual that researchers only called them “feather-like” — this ornithischian (bird-hipped) dinosaur was found with multiple types of scales, some with ribbon-like structures attached, preserved only in two dimensions and lacking any barbs or branching.

Juravenator, a compsognathid first described in 2006, turns out to be the most revealing case. The original researchers noted scale impressions on its tail, and a 2020 paper claimed those impressions weren’t scales at all — a claim the video frames as reflecting a bias toward finding protofeathers. But a detailed 2021 study by Bell and Hendrickx used multiple techniques to document three distinct types of scales on different parts of the tail, along with only a few tiny filaments along the top ridge. Most striking was the thickness of the tail’s soft tissue: six times the height of the tail vertebrae. Applying that same ratio to Sinosauropteryx shows that its “fuzzy” structures would actually sit inside the body wall, not outside it — which fits collagen or muscle fiber far better than external, feather-like integument.

Haolong dongi, an iguanodontid described in 2026, had multiple types of scales, including large plate-like scales along its tail resembling an armadillo or pangolin, plus small spines on its neck and back. The preservation was good enough to see individual cells within the spines: most were tiny, under a tenth of an inch, though some reached nearly two inches. Had the preservation been worse, these spines likely would have been labeled “feathers” or “feather-like” — but because they were so well preserved, they could be identified as keratinous spines instead.

A Testable Prediction

A pattern emerges across these examples: the better a fossil’s preservation, the less feather-like its filaments turn out to be — no barbs, no barbules, no follicles. That leaves several real possibilities for what these structures actually were: collagen, muscle, or tendon fibers; support structures for frills or sails; keratinous spines; or perhaps a rare, hair-like form of integument not seen in reptiles today. As a Christian, Joel doesn’t believe birds evolved from dinosaurs or that birds are dinosaurs — but he’s also clear that this doesn’t rule out God having created some dinosaurs with feathers, since it was God’s choice to make. So far, though, no undisputed branching feathers have been found in fossils of any dinosaur outside Maniraptora, the group of bird-like animals that includes dromaeosaurs, troodontids, and oviraptorosaurs. That leads to a specific, testable prediction: true branching (pennaceous) feathers will only be found on the avian-looking animals within Maniraptora, and none will be found on dinosaurs outside that group — a prediction that new fossil discoveries will keep testing going forward.

Watch on YouTube.