Researchers have discovered what appear to be the remnants of red blood cells and connective tissue in 75 million-year-old dinosaur fossils.
The work could shine a light on long-standing questions about dinosaur physiology, including whether specific species were warm- or cold-blooded.
Chemical analysis revealed similarities between blood cells from fossils and those from living emu.
The work appears in the journal Nature Communications.
Examining part of a fossilised dinosaur claw, the Imperial College London researchers identified tiny ovoid structures with an inner denser core that resembled red blood cells.
And in another fossil fragment, they found fibrous features with a banded structure similar to that seen in modern-day collagen - found in the tendons, skin and ligaments of animals.
It's not the first time such remnants have been found in dinosaur fossils, but co-author Susannah Maidment told BBC News: "All of the previous reports of original components of soft tissues in dinosaur fossils have tended to be in specimens that are really exceptionally preserved - one-offs, really, that require special pleading to explain how they got preserved."
By contrast, the fossils in this study, which have been lying in the London Natural History Museum collections for more than a century, are largely in a poor state of preservation.
"They're very scrappy, individual broken bones. I can't even tell you what dinosaur they come from," said Dr Maidment, who is from Imperial College London.
"If you're finding soft tissues in these kinds of fossils, maybe this kind of preservation might be more common than we realised, and might even be the norm."
The structures appear to be genuine remnants of soft tissue; they are not fossilised.
Using a mass spectrometer, they carried out chemical analysis of the putative collagen protein and the candidate blood cells.
They discovered fragments in the collagen of what look like amino acids - the building blocks of all proteins.
And the chemical profile of the blood cells looked very similar to that obtained from the red blood cells of an emu, which - like all birds - is a direct descendent of dinosaurs.
"There's an extremely well-known relationship within individual vertebrate groups that the smaller the red blood cell, the faster the metabolic rate," said Dr Maidment.
"Animals with fast metabolic rates will tend to be warm blooded, while animals with slower metabolic rates are going to be more cold blooded."
The subject of whether dinosaurs were cold- or warm blooded has preoccupied palaeontologists for decades, because it can provide pointers to the types of lifestyles dinosaurs had.
Were they more bird-like in their behaviour, or more sluggish, like reptiles?
The red blood cells found in this study were small compared with their counterparts in the emu, but the dinosaur cells will have shrunk and curled up over time.
Furthermore, scientists don't yet understand the relationship between red blood cell size and metabolic rate within dinosaurs, so scientists will need a bigger sample from different species of dinosaur to shed useful light on the debate.
However, said Dr Maidment, "if we can find red blood cells in lots of different dinosaurs and measure them, we might be able to start to understand which dinosaurs had fast metabolic rates, which were approaching warm bloodedness, which were truly warm blooded, and which were cold blooded".
Study of the apparent collagen fibres could shed light on the relationships between different dinosaur species.
A technique called collagen fingerprinting is based on the idea that the structure of the collagen molecule is unique to individual animals.
"Most closely related animals will have a more similar collagen structure than more distantly related animals," said Dr Maidment.
"If we could extract some of the collagen... and we could find it in lots of different dinosaurs, it could give us a sense of relatedness within the dinosaur family tree."
Co-author Dr Sergio Bertazzo said: "We still need to do more research to confirm what it is that we are imaging in these dinosaur bone fragments.
"If we can confirm that our initial observations are correct, then this could yield fresh insights into how these creatures once lived and evolved."
Prof Mary Schweitzer from North Carolina State University, who was not involved with the latest study, said she appreciated the caution with which the group interpreted their data.
"All in all, I think that papers like these which present data from multiple lines of investigation, and which are cautious in interpretation do much to advance the field, show that fossils are more than 'just rocks', and open the door to the possibility that materials persist in ancient fossils that were not thought possible only a few years ago," she told BBC News.
Prof Schweitzer added: "They did find amino acids consistent with proteins, but the data they presented do not really identify which proteins; for that they need additional data.
"But it is a great start, and an exciting paper, particularly in showing what happens when you really look at ancient bone and are not bound by the expectation that 'nothing could possibly persist'. If you don't look, you won't find. But if you do, you never know."
On the outside possibility of ever finding DNA in dinosaur remains, Dr Maidment commented: "We haven't found any in our fossils... however, I think it's unwise to say we'll never find any in future."
Follow Paul on Twitter.