Most dinosaur meals left no trace that can be tied to an individual animal. Palaeontologists therefore combine the rare direct evidence of food with anatomy, tooth wear and the plants or animals known from the same geological setting. These clues answer different questions. A jaw may show what an animal could process; stomach contents or a coprolite can record something it actually ate. Neither gives a complete lifetime menu.
The strongest interpretation comes when independent evidence agrees. A study of dinosaur coprolites and gut contents shows both their value and their limits: preserved food can be direct evidence of a meal even when the producer is uncertain.
What teeth and jaws can reveal
Tooth shape provides a starting point. Many theropods had recurved, serrated teeth suited to gripping and cutting flesh. Broad, leaf-shaped teeth or dense dental batteries were adapted for cropping and processing plants. Those features describe mechanical capability, not a complete menu. A sharp tooth does not prove that its owner ate meat exclusively, and a broad tooth cannot identify a particular plant species.
Jaw joints, muscle attachment areas and the arrangement of teeth help researchers test how a bite worked. Computer models can estimate how forces moved through the skull, while fossil bite marks show where teeth contacted bone. These methods narrow the range of plausible feeding behaviour. They usually cannot distinguish a hunted animal from a carcass found after death, or identify every food eaten by the dinosaur that left a mark.
Wear records repeated contact
Scratches and microscopic wear on tooth surfaces reflect contact with food and grit. Their direction, depth and distribution can be compared across many teeth. In hadrosaurs and ceratopsians, dental batteries brought multiple teeth into a continuously renewed grinding or slicing surface. Wear and replacement histology explain how these systems functioned, but they do not name the leaves or stems that passed through them.
Microwear can record a relatively short interval before death. Sediment grains may leave marks that resemble those made by tough plants, and different foods can produce overlapping patterns. Researchers therefore compare wear with jaw movement, tooth anatomy, gut contents and local plant fossils rather than treating a scratch pattern as a menu by itself.
Direct remains of food
Food preserved inside a body outline or digestive region can link plant or animal remains to one individual. Such finds are uncommon, and they usually record a final meal rather than a normal diet over an entire season or lifetime. A 2025 study of gut contents in the sauropod Diamantinasaurus identified remains from conifers, seed ferns and flowering plants. The specimen supports a broad diet in that juvenile animal, not a universal menu for sauropods.
Coprolites preserve material after digestion. Plant fragments, phytoliths, bone and other resistant remains can identify parts of a meal. A large Late Cretaceous coprolite containing fragmented bone is consistent with a large theropod, probably a tyrannosaurid, but the producer was not found inside the specimen. Its contents are direct evidence of digested bone; the animal's identity remains an inference.
Several clues are better than one
Teeth, jaw mechanics, wear, gut contents, coprolites, bite marks and the surrounding flora each describe a different part of feeding. Agreement between them can support a stronger conclusion than any one clue alone. Disagreement can reveal that a group included different feeding strategies, that diet changed with age, or simply that the fossil record is incomplete.
That is why palaeontologists describe a supported range rather than inventing a precise menu. The evidence for plant-eating dinosaurs includes several different tooth and digestive strategies, while a Jurassic food web helps place those animals among plants, predators and smaller consumers without implying that every possible interaction is known.
Frequently asked questions
Do sharp teeth prove that a dinosaur was a predator?
No. Tooth shape shows how a tooth could work, but the animal's diet is inferred from several kinds of evidence, including wear, jaw mechanics and direct food remains.
Can a coprolite identify the dinosaur that produced it?
Sometimes its size, contents and geological context narrow the possibilities. Unless it is directly associated with an animal, the producer may remain uncertain.
Do fossil gut contents show an animal's usual diet?
They record food eaten shortly before death. A single meal cannot establish everything an animal ate through its life or in different seasons.
How do researchers tell whether a dinosaur was an omnivore?
They compare teeth, jaw function, wear and direct food remains. A mixed diet is most persuasive when several independent clues support it.