Traces of dinosaur blood survived 115 million years inside a leg bone from Brazil

Inside the leg bones of a meat-eating dinosaur from northeastern Brazil, geochemists have found tiny rounded bodies packed into the channels that once carried its blood vessels, along with chemistry that ties them to blood. The bodies carry heme, the iron-bearing molecule at the heart of hemoglobin, the protein red blood cells use to carry oxygen.

If the identification holds, it is the first time anyone has chemically matched heme to fossil red blood cells still sitting in place inside a bone. That would push back the limit on how long a fragile biological molecule can last. The animal, Aratasaurus museunacionali, lived in the Early Cretaceous, more than a hundred million years ago, and its bones have already outlasted one modern disaster: the fire that gutted Brazil’s National Museum.

The study, led by the organic geochemist Kliti Grice of Curtin University in Australia and published in the Proceedings of the National Academy of Sciences, describes the bodies as mineralized red blood cells preserved for 115 million years. “To detect intact heme derived molecules in 115-million-year-old red blood cells is astonishing: this really is a landmark discovery,” Grice said in a statement from the university.

What the team did not find matters as much as what it did. There is no dinosaur DNA here and no intact hemoglobin protein, only the heme ring and a chemically altered version of it. Grice told Reuters the result is not a “vial of dinosaur blood.” “What we are seeing is a mineral and molecular echo of ancient biology, and that is scientifically much more interesting,” she said.

The dinosaur born from a fire

Aratasaurus is known from a single partial right leg, found in 2008 in the Mina Pedra Branca plaster mine near Santana do Cariri, in Ceará state, in rocks of the Romualdo Formation. The rest of the skeleton was broken up and lost during mining. The leg went first to the Plácido Cidade Nuvens Museum of Paleontology, where thin slices were cut to study the bone tissue, and in 2016 it moved to the National Museum in Rio de Janeiro.

On 2 September 2018, fire swept through that museum, the country’s oldest scientific institution, and destroyed an estimated 92% of its 20 million items. The area where the leg was stored was not touched. When Juliana Sayão and colleagues formally named the animal in 2020, they joined the Tupi words ara, “born”, and atá, “fire”, and dedicated the species to the museum: the dinosaur born from the National Museum fire. Sayão and Alexander Kellner, a paleontologist at the museum, are co-authors of the new paper.

Growth marks in the bone suggest the animal was about four years old when it died, a young theropod roughly 3 meters long and still growing. It belonged to the coelurosaurs, the branch of two-legged predators that later produced the tyrannosaurs and, eventually, birds.

How they did it

Grice was not hunting for dinosaur blood. Her lab studies how much of an organism’s original chemistry can survive fossilization. Under the microscope, the vascular canals of the Aratasaurus leg turned out to be filled with small ellipsoidal bodies about 23 micrometers long and 20 wide, roughly three times the diameter of a human red blood cell.

Shape alone proves nothing, because minerals and microbes both make convincing lookalikes of cells. So the team combined microscopy, mineral analysis and molecular geochemistry. Extracts of the bone went through high-performance liquid chromatography coupled to mass spectrometry, a technique that separates a mixture into its components and weighs each one. It turned up heme B, the most common form of heme in hemoglobin, and an altered relative called 3,8-diethyl heme B, which was about 90 times more abundant. Most of the heme, in other words, had been chemically changed during fossilization.

Spectroscopy then found the signature of porphyrin, the ring structure at the core of heme, inside the cell-like bodies themselves, and elemental maps showed iron, nitrogen and carbon in the same spots. “Critically, we investigated the chemistry in situ, so we could relate the chemical signals to the microscopic structures in their physical context within the fossil,” Grice told Live Science. “That is much stronger than simply extracting molecules from a piece of fossil and assuming where they originated.”

To check for contamination, the researchers analyzed the surrounding rock the same way. It held no heme B, only the signatures of chlorophyll and bacteriochlorophyll, pigments of the photosynthetic microbes that formed the mat in which the leg was buried. The bone also contained hydroxyproline, an amino acid typical of collagen, the protein that gives bone its toughness.

Why blood chemistry could last this long

The cell-like bodies are coated in hausmannite, a manganese oxide mineral. The team proposes that it formed early and worked like a time capsule, sealing the structures before they decayed. They also think the carcass was buried quickly in a low-oxygen setting covered by microbial mats, which would have slowed normal decomposition.

“We think early mineralisation helped stabilise or replicate microscopic biological structures before they were completely destroyed, while aspects of the associated chemistry also survived in altered form,” Grice told Live Science. In the university statement she went further: “These findings challenge longstanding assumptions that certain biomolecules such as haemoglobin cannot survive deep time.”

That assumption has been under pressure for decades. Researchers reported heme-containing compounds from a Tyrannosaurus rex bone in the late 1990s, and a 2015 study described possible red blood cells and collagen-like fibers in 75-million-year-old dinosaur specimens. Both claims were fiercely debated. The new paper’s difference is that it ties the chemistry to specific structures inside the bone rather than to a powder ground from it.

What it doesn’t settle

Not everyone accepts that these are blood cells. Mary Schweitzer, a molecular paleontologist at North Carolina State University who has reported blood vessels and soft tissue preserved in dinosaur fossils, was not involved in the study. “I am in agreement that endogenous molecules can persist over geological time in some form, and that there is much these can teach us,” she told Live Science. “However, I am not convinced that the images they figure are remnant RBCs for several reasons.”

Her main objection is shape. Mammals strip the nucleus out of their red blood cells, but birds and reptiles keep it, which gives each cell a raised bulge in the middle. The Aratasaurus bodies are concave, with no bulge, and smaller than most red cells of non-mammals, “so morphological similarity is not met,” Schweitzer said. The cells might have been squashed or ruptured before they mineralized, or dinosaur red cells may simply have differed from those of their living relatives, but neither idea has been tested. Schweitzer found the heme and collagen chemistry more convincing than the identification of the cells.

Contamination is the second open door. The paper itself acknowledges that heme-producing microbes cannot be completely ruled out. A study of Centrosaurus bones in 2019 found recent organic material and thriving microbial communities living inside the fossil, a reminder that buried bone can become a habitat with its own chemistry. The clean result from the surrounding rock argues against an outside source, but it cannot exclude microbes that lived inside the bone.

The rest of the limits are about scale. This is one leg from one animal at one site. Hydroxyproline is a building block, not a collagen protein or a sequence, and nothing in the study shows the protein part of hemoglobin. The hausmannite time capsule is a proposed pathway, not an experiment anyone has reproduced in the lab.

Common questions about dinosaur blood

Did scientists find real dinosaur blood?

Not liquid blood. They found structures that look like red blood cells, turned to mineral, and heme molecules associated with them. Most of that heme had been chemically altered over 115 million years.

Can dinosaur DNA be recovered from this fossil?

No. The study reports no DNA, and Grice has stressed that the aim was never to find it. The oldest DNA ever recovered comes from Greenland sediments about two million years old, far younger than any dinosaur.

What is heme?

Heme is a ring-shaped molecule with an iron atom at its center. Each hemoglobin molecule holds four of them, and they are what binds oxygen in the blood. Heme B is the most common form.

What kind of dinosaur was Aratasaurus?

It was an early coelurosaur, a lightly built two-legged predator about 3 meters long, from the Early Cretaceous of northeastern Brazil. It is known only from a partial right leg kept by the National Museum of Brazil.

What comes next

The paper appeared online in the Proceedings of the National Academy of Sciences on 2 November 2026. The team, which includes researchers from Curtin, the Japan Agency for Marine-Earth Science and Technology, the National Museum, the Regional University of Cariri and Johns Hopkins University, wants to use the same in-place approach to study the evolution of blood, metabolism and activity levels in extinct animals.

That work may not need new digs. “The really exciting implication is that extraordinary molecular information may already be sitting in museum collections around the world, waiting for us to ask new questions of old fossils,” Grice told The Guardian. The first fossil to make that case was sitting in one when the building burned.

Reference: Grice et al., “Heme signatures in mineralized red blood cells within vascular canals of a 115-million-year-old dinosaur bone,” Proceedings of the National Academy of Sciences, 2026. DOI: 10.1073/pnas.2614248123

Tags: dinosaurs, paleontology, Curtin University, Aratasaurus, hemoglobin, National Museum of Brazil