When Jurassic Park's velociraptors systematically tested electric fences and coordinated ambush attacks, it felt thrillingly plausible. But how much of that was science, and how much was cinema? The question of dinosaur intelligence is more nuanced — and more interesting — than either Hollywood or the old 'walnut-brained' stereotype suggests.
How We Measure Fossil Intelligence
The brain itself almost never fossilises. What palaeontologists study instead are endocasts — natural or artificial casts of the interior of the skull, which preserves the rough shape of the brain. From endocasts, researchers can calculate brain volume and, crucially, the Encephalisation Quotient (EQ): the ratio of actual brain size to the expected brain size for an animal of that body weight.
An EQ of 1.0 means average intelligence for that body size; higher numbers suggest greater cognitive capacity relative to body mass. Modern crows and ravens score around 2.0–2.5. Humans score approximately 7.5. Most large dinosaurs scored between 0.1 and 0.4 — considerably below average for modern reptiles.
Stegosaurus is the classic example of a low-EQ dinosaur: its brain weighed approximately 80 grams in a body of 5 tonnes — an EQ of roughly 0.06. It likely had extremely limited cognitive processing, probably comparable to a modern crocodilian.
The Troodon Problem
Then there is Troodon formosus. This small Cretaceous theropod from North America had an EQ estimated at around 0.9–1.0 — comparable to some modern birds and significantly higher than any other known non-avian dinosaur. Its forward-facing eyes gave it stereoscopic depth perception. Its hands were grasping. Its body plan was highly bird-like in multiple respects.
In 1982, Canadian palaeontologist Dale Russell and artist Ron Séguin conducted a famous thought experiment: if Troodon had not gone extinct, what might 50 million years of continued evolution have produced? Their answer was the 'Dinosauroid' — a bipedal, large-brained, humanoid creature. Though widely criticised as too anthropocentric, it captured imaginations and raised a genuinely serious question: is large-brained bipedal intelligence an evolutionary attractor that multiple lineages might converge upon?
Pack Hunting: Fact or Fiction?
The idea that Velociraptor or Deinonychus hunted in coordinated packs — like wolves — was widely popularised after the 1969 discovery of a Deinonychus quarry containing multiple individuals alongside an adult Tenontosaurus. John Ostrom, who described the find, suggested cooperative hunting.
More recent analysis has complicated this picture considerably. The multiple Deinonychus individuals may represent not a coordinated pack, but individual predators drawn to the same carcass, possibly killing each other in competition. Komodo dragons aggregate around carcasses in a similar way — and nobody would call that cooperative hunting. The evidence for true coordinated pack behaviour in any non-avian dinosaur remains circumstantial.
Parental Behaviour as Evidence of Cognition
One of the most reliable indicators of cognitive capacity is parental care behaviour. Here, several dinosaur groups show genuinely impressive evidence. Maiasaura nests show juveniles at multiple developmental stages in the same nest, with parents clearly returning to feed them. Oviraptor specimens are found brooding their eggs in a posture identical to modern birds. These behaviours require memory, recognition of offspring, and the ability to modify behaviour based on environmental feedback — all markers of meaningful cognitive capacity.
The honest answer to whether dinosaurs were intelligent is: it depends entirely on the species. Most large herbivores were probably operating at crocodile-level cognition. Small theropods — particularly the coelurosaurs that gave rise to birds — were arguably already bird-brained in the most complimentary sense. And in at least one lineage, the cognitive trajectory that eventually produced the crow, the parrot, and the human brain was already well underway.