Dinosaurs had hearing systems, but the soft parts that shaped sound almost never fossilise. Palaeontologists study the bony labyrinth of the inner ear, the middle-ear region, rare laryngeal bones and structures such as the long nasal passages of Parasaurolophus. These clues constrain what animals could hear or produce; they do not preserve a recording of a call.
What the inner ear preserves
The inner ear sits inside the skull. Computed tomography can make a three-dimensional model of the surrounding bone without cutting into a fossil. The semicircular canals are mainly associated with balance and head movement, while the cochlear region is relevant to hearing. Their shapes help researchers compare extinct animals with living birds and crocodilians.
The length of the bony cochlear canal has been compared statistically with hearing traits in living archosaurs. It is not a direct scale of hearing range. Body size, evolutionary relationships and the missing soft tissues all affect the comparison. A study in 2025 cautioned against using relative canal length by itself to rank dinosaur hearing ability or infer behaviour.
A thin bone called the stapes transmitted vibration through the middle ear toward the inner ear. The eardrum and other soft structures rarely survive, so sensitivity and exact frequency limits must be estimated indirectly. A fossil skull therefore offers a partial anatomical record, not a complete acoustic instrument.
Balance and hearing are related but distinct
The semicircular canals lie near the hearing structures, but they chiefly inform questions about balance and head motion. Their shape can contribute to a model of how an animal moved its head. A relatively large canal does not automatically mean exceptionally sharp hearing, a particular hunting method or a specific social behaviour.
Could dinosaurs roar?
There is no direct evidence that non-avian dinosaurs produced a lion-like roar. Birds make sound with a syrinx near the base of the trachea, while crocodilians use a larynx. These living archosaurs show that complex signals can be produced in different ways, but neither is an exact template for every extinct dinosaur.
Comparisons among living archosaurs allow quieter, closed-mouth humming or cooing signals as possibilities for some dinosaurs. They do not prove that every group used those sounds. The distinctions are explored further in what dinosaur voices may have sounded like, where possible sound models are separated from evidence of an actual call.
A rare larynx and a resonating crest
Fossilised, partly ossified laryngeal elements have been described in the ankylosaur Pinacosaurus. Their proportions resemble structures that help living birds modify sound. This rare find adds anatomical evidence about the sound-producing system, but it does not reveal the animal's exact pitch, loudness or timbre.
The long crest of Parasaurolophus contains curved nasal passages. Acoustic models indicate that the passages could resonate at low frequencies, making a role in communication plausible alongside visual display. A model estimates the resonant properties of a reconstructed shape; it does not recover one definitive “Parasaurolophus call”. Results depend on the canal reconstruction, the sound source and soft tissues that did not fossilise.
Signals do not prove a social system
Calls could have helped animals recognise one another, warn of danger, attract mates or keep contact between adults and young. Eggs, nests, trackways and bonebeds can preserve evidence relevant to social behaviour, but the hearing apparatus alone cannot prove herding, coordinated hunting or migration calls. Signals might also have been visual, using posture, movement, colour or a conspicuous crest.
Each fossil line answers a different question. CT models show bony spaces; a rare larynx constrains sound production; an acoustic model estimates resonance. Bringing them together can make communication plausible while leaving the actual sound and its purpose uncertain. The brain and sensory evidence described in dinosaur brain studies provides a related, but separate, line of inquiry.
Frequently asked questions
Can scientists know exactly what a dinosaur sounded like?
No. Fossils can constrain anatomy and model possible resonances, but the soft tissues and sound recording needed to recover an exact voice are missing.
Does a long cochlear canal prove that a dinosaur heard better?
Not by itself. Canal length has statistical relationships with hearing traits in living animals, but body size, ancestry and soft tissues affect the comparison.
Did Parasaurolophus use its crest to make a call?
Its curved nasal passages could resonate at low frequencies in acoustic models, so sound production is plausible. The models do not establish one exact call or prove that sound was the crest's only function.
Does a fossilised larynx reveal a dinosaur's voice?
A rare larynx can constrain the anatomy used to produce sound, as in Pinacosaurus, but it does not preserve pitch, loudness or timbre.