Dinosaur fights: what fossils can and cannot tell us

“Which dinosaur would win?” is a familiar question, but fossils rarely provide a scoreboard. They can preserve bites, fractures, healed wounds, stomach contents and, in exceptional cases, two animals buried together in close contact. These traces help reconstruct real interactions. They do not reveal how every individual behaved or settle imagined contests between species that never met.

A meaningful comparison starts with geography and time. Animals must have lived in overlapping places and intervals before an encounter was even possible. Tyrannosaurus and Triceratops shared parts of western North America near the end of the Cretaceous, while Spinosaurus lived earlier and in a different region of North Africa. A matchup between the latter and Tyrannosaurus is a thought experiment, not a fossil-based event.

The rare direct record of an encounter

The most famous example is the “Fighting Dinosaurs” specimen from Mongolia. It preserves a Velociraptor and a Protoceratops in close contact. The raptor’s foot claw lies near the herbivore’s neck, and the Protoceratops jaws hold the predator’s forelimb. The fossils show a direct interaction between these animals; the exact sequence and cause of death are harder to reconstruct.

The specimen was found in the Upper Cretaceous Djadokhta Formation. Rapid burial is often proposed to explain how the animals remained together, but the precise sedimentary event is not witnessed by the skeletons. Later work on additional remains has also supported a predator–prey relationship between Velociraptor and neoceratopsians. That strengthens the ecological link without making every detail of the famous encounter certain.

Other paired fossils can preserve possible interactions, but close association alone is not enough. Animals may have accumulated together after death, or remains may have been moved by water. Scientists examine articulation, sediment, injury and the position of bones before deciding whether an interaction occurred while the animals were alive. A fossil pose is not automatically a snapshot of behaviour.

Bites, wounds and survival

Tooth marks on bone can indicate feeding, attack or postmortem scavenging. Their shape and spacing may help identify the likely biter, but the mark’s location and state of healing matter. A wound with remodelling bone shows that the animal survived for some time after injury; a sharp, unhealed break may have happened near death or after it. Neither automatically tells us who initiated a fight.

Some Triceratops skulls preserve healed injuries that have been discussed as evidence of combat between individuals. A statistical study compared lesions on Triceratops and Centrosaurus and found patterns consistent with horn use in intraspecific interactions. The result supports a combat hypothesis for some injuries, but it does not show that every skull lesion was caused by a horn or that fights were the only use of horns.

Predator bites also appear on dinosaur bones. A hadrosaur vertebra with a tyrannosaur tooth lodged in it shows healing around the injury, providing strong evidence that a large animal survived an attack. Such fossils reveal a specific event and a likely predator. They cannot establish a routine hunting success rate or tell us whether a predator usually hunted or scavenged.

Researchers distinguish marks made while an animal was alive from those made after death by examining healing, bone response, fracture surfaces and sedimentary context. The distinction is not always clear. A tooth can score a carcass after death, and a partially healed injury may have occurred well before the final burial. Each specimen has its own taphonomic history.

Why anatomy does not decide a winner

Skull shape, teeth, horns, armour, body mass and limb proportions all matter, but their effects depend on the situation. A large bite could cause severe damage if it landed; turning, balance, fatigue, footing and the ability to avoid an attack also mattered. A skull model estimates how stresses might travel through bone under chosen assumptions. It does not recreate every muscle, movement or decision in a living animal.

Body-size estimates also vary. They depend on how complete a skeleton is, which living animals are used for comparison and how much soft tissue is added to a model. Comparing one species’ largest known adult with another species’ average or juvenile creates a misleading contest. The same caution applies to bite-force estimates and weapons: a maximum theoretical force is not a guaranteed result in an encounter.

Fossil weapons also had costs and constraints. A horn could injure an opponent but expose its owner to a broken bone; heavy armour adds mass; a long tail club needs room to swing. These mechanical considerations can help test a proposed function, yet they cannot tell us how often a structure was used in combat or which target an animal chose.

There is also a difference between evidence that an animal was attacked and evidence that it was killed by that attack. Healed bone is strong evidence of survival; a fatal injury is harder to establish because death may leave no distinctive mark. A carcass could attract a predator after an animal died from illness, starvation or another cause. Feeding traces alone do not identify the event that caused death.

The well-known paired skeleton from Mongolia is exceptional partly because two animals are preserved together. Most encounters left no fossil trace, and most fossils that do preserve a bite record only a small part of an interaction. A single specimen cannot tell us how common the behaviour was across a species, whether adults and juveniles behaved alike or whether the same interaction happened regularly.

Models can narrow possibilities without supplying a play-by-play. A biomechanical analysis may show that a skull could withstand a certain loading direction, or that a tail could deliver a forceful swing. Those results depend on the reconstructed muscles, posture and loading assumptions. They describe mechanical capacity under a model, not a guaranteed tactic or outcome.

What plausible comparisons can say

When two species shared a time and place, fossils and anatomy can outline possible interactions. Tyrannosaurus and Triceratops lived in the same Late Cretaceous ecosystems, and regional fossils include both predatory bite marks and healed injuries. This makes predator–prey contact and defensive encounters plausible. The fossils do not show that every meeting became a fight or name a universal winner.

Allosaurus and Stegosaurus co-occurred in some Jurassic settings. The predator’s teeth and skull and the herbivore’s plates and tail spikes suggest a dangerous interaction if one occurred. A famous “duel” between them is still a reconstruction. The two animals did not leave a single paired fossil that proves how a particular confrontation unfolded.

Other popular matchups fail a basic test of coexistence. Spinosaurus and Tyrannosaurus came from different continents and geological intervals, while Velociraptor and Triceratops were not contemporaneous neighbours. Anatomical comparisons can still be useful as biomechanics exercises, but they should not be presented as probable prehistoric encounters.

Direct evidence is most useful when it replaces a broad claim with a specific one. The Velociraptor and Protoceratops specimen documents contact; healed injuries document survival after trauma; tooth marks document feeding or attack under the right conditions. These clues make dinosaur lives less abstract while keeping the unknown parts visible.

The word “fight” itself can hide different behaviours. A defensive strike, a bite used to restrain prey, competition at a carcass and a contest between members of one species are not the same event. Fossils may support one interpretation more strongly than another, but an elaborate narrative about motive usually goes beyond what bone can preserve.

From fossil evidence to responsible reconstruction

A scientific account separates what is preserved from what is inferred. A fossil may show two skeletons touching; their posture may suggest combat; sediment may indicate rapid burial; the exact trigger and sequence may remain unknown. Artwork can show one plausible scenario, but its caption should say when the scene is interpretive.

That approach still allows vivid stories. Dinosaur fights were possible, sometimes lethal and occasionally preserved in extraordinary detail. The strongest conclusion is not a ranking of champions, but a record of which animals shared ecosystems, what injuries they sustained and how the evidence constrains our reconstructions.