There is no single answer to the question “what colour were dinosaurs?” Most fossils preserve bones, not skin pigments, and colour is rarely recoverable. In a small number of exceptionally preserved specimens, microscopic structures called melanosomes and patches of skin or feathers retain clues to pigment. These findings support specific colour patterns for particular animals. They do not justify painting every dinosaur green, grey or brown.
How pigment evidence survives
Melanosomes are structures inside pigment-producing cells. In living animals, their shape and chemistry contribute to colours such as black, grey, brown and reddish tones. Researchers compare fossil melanosomes with those in modern birds and other animals, using microscopy and chemical analyses. A 2010 study identified melanosomes in the filament-like covering of Sinosauropteryx and inferred chestnut to reddish-brown tones in dark bands along its tail.
This is an inference built from a chain of evidence: the structures must be correctly identified, their distribution must reflect the original integument, and comparisons with living animals must be relevant. It is stronger than choosing a colour from artistic intuition, but it is not a perfect colour photograph. Preservation, mineral replacement and the limits of modern comparisons all affect the result.
Patterns can reveal more than a colour name
In Sinosauropteryx, the placement of pigment-bearing feathers supports a striped tail and a darker back than underside, a pattern called countershading. A 2017 study compared this pattern with modern animals and proposed that the distribution could inform ideas about habitat and light environment. Such ecological interpretations are one step further from the fossil: the colour pattern is inferred from preserved structures, while the habitat explanation is a model that can be tested against geology and other evidence.
Preserved skin in the ceratopsian Psittacosaurus provides a different kind of evidence. Its body retains scales and a countershaded pattern, with darker upper surfaces and lighter undersides. Because this is not a feathered theropod, it shows that colour evidence is not confined to one narrow dinosaur group. Even so, it represents one species and one well-preserved individual, not a template for all ceratopsians.
Why the method needs safeguards
Not every melanosome-like body in a fossil came from skin or feathers. A 2018 study showed that internal tissues can contain melanosomes too, and that decay may move them around a carcass. If a fossil sample is taken from the wrong region or the tissue source is uncertain, a colour reconstruction can be biased. Researchers therefore consider where the structures occur, whether surrounding anatomy supports an integumentary origin, and how the sample was preserved.
Other colours may be made by pigments or structures that do not fossilise in the same way. Iridescence, structural blues and greens, carotenoid-based hues and patterns in bare skin may leave little direct trace. A fossil that preserves black or reddish melanin clues may still tell us nothing about a display patch or the colour of the eyes. The absence of colour evidence is not evidence that the animal was dull-coloured.
What artists can responsibly show
Scientific illustrations combine layers with different confidence. Bone shape and feather impressions may be directly preserved; pigment distribution may be measured; the body outline and soft tissues require reconstruction; unpreserved colours may be chosen for visual clarity. A caption should say which details are supported and which are hypothetical. A convincing image can help readers picture an animal, but visual confidence should not be mistaken for fossil evidence.
Colour plates and digital models can also make uncertainty explicit by showing alternative reconstructions. One version may display only the pigment pattern supported by a specimen; another may explore a plausible but untested choice for the rest of the body. This is useful in museums and publications because readers can see where evidence ends. It also prevents a frequently repeated illustration from becoming mistaken for a direct observation.
As new fossils and analytical methods appear, specific reconstructions can change. Researchers may refine the location of a pattern, revise an animal's classification or distinguish skin-derived melanosomes from those in other tissues. The most defensible summary is therefore modest: some dinosaurs had recoverable patterns, and some may have been vividly coloured, but the palette of most species remains unknown.
See how feather anatomy informs reconstruction in the evolution of dinosaur feathers and compare preserved skin with the Psittacosaurus profile.
Frequently asked questions
Do we know what colour T. rex was?
No reliable pigment pattern has been established for Tyrannosaurus. Popular colour schemes are artistic choices unless linked to direct fossil evidence.
How can melanosomes reveal ancient colour?
Researchers identify pigment structures in fossil integument and compare their form and chemistry with living animals. The result is an evidence-based inference with preservation limits.
Were all dinosaurs green or grey?
There is no evidence for one shared dinosaur palette. A few fossils preserve patterns, while the colours of most species are unknown.
Can colour patterns reveal habitat?
Sometimes they can support hypotheses, such as countershading for camouflage, but the habitat interpretation is less direct than the fossil pigment evidence.