DNA can survive in fossils, but only under narrow conditions and never indefinitely. The molecule breaks apart over time through chemical reactions that chop it into ever-shorter fragments, so what researchers recover is always damaged and incomplete. The oldest authenticated DNA comes from sediment in northern Greenland dated to roughly two million years ago, while the oldest genome sequenced from an identifiable organism belongs to mammoth teeth over a million years old, preserved in Siberian permafrost. Beyond a certain age and temperature history, no usable genetic material remains, no matter how well a fossil looks on the outside.
How DNA Falls Apart in Fossils
The moment an organism dies, its DNA begins to degrade. Enzymes from the body’s own cells and from invading microbes start the process, but over archaeological and geological timescales the main culprits are purely chemical. Water molecules attack the bonds that hold the DNA backbone together, snapping it into shorter and shorter pieces. A particularly important reaction called depurination strips away certain chemical bases, creating weak points where the strand breaks. On top of that, cytosine residues gradually lose an amino group and turn into uracil, a change that clusters toward the ends of surviving fragments and acts as a chemical fingerprint researchers use to confirm that DNA is genuinely old rather than modern contamination.1PubMed Central. Ancient DNA damage
A study of 158 dated moa bones from New Zealand estimated that a roughly 240-base-pair stretch of mitochondrial DNA has a half-life of about 521 years under those burial conditions. That rate turned out to be almost 400 times slower than what lab experiments on naked DNA in solution would predict, probably because bone mineral shields the molecule somewhat.2PubMed Central. The half-life of DNA in bone: measuring decay kinetics in 158 dated fossils Even so, a half-life of a few centuries means that after tens of thousands of years the surviving fragments are tiny, often under 50 base pairs, and after hundreds of thousands of years you are working with molecular confetti. The rate depends heavily on temperature: warm burial sites accelerate every one of these reactions, while freezing slows them dramatically.
Why Cold Matters So Much
Almost every record-breaking ancient DNA result has come from permafrost or high-latitude sites. The reason is straightforward: the chemical reactions that destroy DNA are temperature-dependent. A bone sitting in tropical soil at 20°C loses its DNA orders of magnitude faster than one locked in permafrost at minus 10°C. Researchers working in the tropics have struggled to recover usable ancient DNA at all, because factors like heat and water infiltration chew through the molecule before it can be preserved.3PubMed Central. Ancient DNA and the tropics: a rodent’s tale Permafrost, by contrast, provides quantitatively lower degradation rates than almost any other natural setting.4Nucleic Acids Research. New insights from old bones: DNA preservation and degradation in permafrost preserved mammoth remains
Temperature is not the whole story, though. Aridity helps because water drives hydrolysis. Stable pH matters. And the specific skeletal element makes a surprising difference. The petrous bone, a dense pyramid of bone encasing the inner ear, holds at least three times more DNA-containing cells than most other bones, giving it a head start that translates into better preservation over time.5PubMed Central. The petrous bone contains high concentrations of osteocytes: One possible reason why ancient DNA is better preserved in this bone Direct comparisons have shown that petrous bone and tooth cementum both vastly outperform other skeletal elements, sometimes by a factor of ten or more in terms of the proportion of DNA that actually belongs to the organism rather than to soil microbes.6PLOS ONE. Comparing Ancient DNA Preservation in Petrous Bone and Tooth Cementum If you have a choice of which bone fragment to sample, those two are almost always where you start.
The Current Records
The oldest DNA ever authenticated comes from environmental DNA, meaning genetic fragments shed into sediment by organisms rather than recovered from their bones. A 2022 study pulled DNA from clay-rich sediments at Kap København in northern Greenland, dating to roughly two million years ago. The sequences revealed an entire ecosystem: mastodon relatives, reindeer, hares, plants, and microbes living in a boreal forest that no longer exists at that latitude.7Nature. A 2-million-year-old ecosystem in Greenland uncovered by environmental DNA The DNA survived in part because it had bound tightly to clay and quartz minerals, protecting it from water.
For DNA extracted from identifiable skeletal remains, the record belongs to mammoth teeth recovered from eastern Siberian permafrost. Researchers sequenced genome-wide data from three specimens, two of which are over a million years old.8Nature. Million-year-old DNA sheds light on the genomic history of mammoths The sequences were highly fragmented and required heavy computational processing, but they were sufficient to reveal that one of the specimens belonged to a previously unknown mammoth lineage that later contributed ancestry to the Columbian mammoth of North America. That finding would have been invisible without the ancient DNA.
The Jurassic Park Myth
In the early 1990s, a few research groups reported extracting DNA from insects preserved in amber tens of millions of years old, a finding that conveniently aligned with the plot of a blockbuster film. Those claims did not hold up. When independent labs tried to reproduce the results under strict contamination controls, they found no authentic insect DNA in amber or copal samples at all.9PubMed Central. Problems of reproducibility–does geologically ancient DNA survive in amber-preserved insects? The original sequences were almost certainly modern contamination that slipped through the less rigorous protocols of the time.
Amber might look like the perfect preservative, but it does not stop the chemistry of degradation the way permafrost does. Amber-entombed insects are typically millions of years old and have been stored at moderate temperatures for all of that time. Given what we know about DNA decay rates, no readable DNA should survive under those conditions, and repeated attempts to find it have confirmed that it does not. Dinosaur fossils are in an even worse position: they are tens of millions of years old, have been mineralized, and were never frozen. There is no credible evidence that dinosaur DNA exists in any fossil, full stop. Claims to the contrary surface periodically and invariably turn out to be contamination or misidentified proteins.
Getting DNA Out of Ancient Samples
Extracting ancient DNA requires methods quite different from what a modern genetics lab uses, because the molecules are so short and so damaged. Standard double-stranded library preparation methods lose many of the tiniest fragments, which are often the majority of what is present. Single-stranded library preparation methods were developed specifically to capture these scraps. They work by converting each strand independently into a form that a sequencing machine can read, which roughly doubles the amount of usable data from a given extract.10PubMed. A Method for Single-Stranded Ancient DNA Library Preparation Several competing protocols now exist, each refining the chemistry to reduce cost, increase throughput, or minimize biases in which fragments get captured.11Nucleic Acids Research. Single-stranded DNA library preparation from highly degraded DNA using T4 DNA ligase
Once sequences come off the machine, computational tools sort the ancient from the modern. Authentic ancient DNA carries characteristic damage patterns, particularly an excess of apparent C-to-T changes near fragment ends caused by cytosine deamination. Software tools quantify this damage signature across assembled sequences and flag material that does not show it as likely contamination.12PubMed Central. PyDamage: automated ancient damage identification and estimation for contigs in ancient DNA de novo assembly This is one of the field’s neatest tricks: the very chemical damage that makes ancient DNA hard to work with also serves as proof that it is genuinely old.
Contamination Is the Persistent Enemy
The biggest practical challenge in ancient DNA work is not getting enough DNA out of a sample; it is making sure the DNA you get actually belongs to the organism you are studying. Every person who has ever handled the specimen, every bacterium in the soil, and every stray skin cell drifting through the lab contributes foreign DNA. In a typical ancient bone sample, the vast majority of recovered sequences come from environmental microbes rather than from the animal itself. Pre-treatment steps help. Exposing bone powder to phosphate solutions can strip away about two-thirds of microbial DNA while removing only about a third of the endogenous DNA, effectively doubling the proportion of useful sequences.13PubMed. Reducing microbial and human contamination in DNA extractions from ancient bones and teeth
Different sample types bring different contamination challenges. Dental calculus, the hardite buildup on teeth, preserves DNA well but is overwhelmingly microbial in origin. The host organism’s DNA typically makes up less than half a percent of the total.14Scientific Reports. Differential preservation of endogenous human and microbial DNA in dental calculus and dentin For calculus specifically, decontamination protocols using chemical washes or UV irradiation before extraction can shift the balance, reducing environmental microbe DNA and increasing the proportion of oral-origin sequences.15PubMed Central. Effectiveness of decontamination protocols when analyzing ancient DNA preserved in dental calculus Modern labs also operate under strict clean-room conditions, with separate pre-PCR areas, protective suits, and extensive blank controls run alongside every real sample.
DNA From Dirt
You do not always need a bone or a tooth. Sedimentary ancient DNA, recovered from cave floors or lake beds, has opened a new front in paleogenetics. Organisms shed DNA constantly through skin cells, urine, feces, and decomposition, and some of that genetic material binds to mineral grains in the surrounding soil. Researchers have used targeted capture methods to pull hominin mitochondrial DNA from cave sediments at sites across Eurasia, detecting Neanderthal DNA in eight archaeological layers from four caves and Denisovan DNA in a deep layer of Denisova Cave, including at sites and in layers where no skeletal remains had been found at all.16PubMed. Neandertal and Denisovan DNA from Pleistocene sediments
Lake sediments offer a different window. Researchers have successfully sequenced plant and animal DNA from Holocene lake sediments in North America, using bulk sediment as well as identifiable plant fragments like seeds and leaf pieces.17PubMed Central. Ancient DNA from lake sediments: bridging the gap between paleoecology and genetics Sedimentary DNA will never give you a complete genome the way a well-preserved bone can, but it is transforming the field by making genetic data available from places and time periods where no fossils survive.
What Ancient Genomes Have Revealed
The practical payoff of fossil DNA is enormous. One bone fragment from Denisova Cave, barely large enough to notice, turned out to belong to a girl whose mother was Neanderthal and whose father was Denisovan. The fact that a first-generation hybrid was found among the tiny handful of archaic human specimens ever sequenced suggests that interbreeding between these groups was common wherever they overlapped.18Nature. The genome of the offspring of a Neanderthal mother and a Denisovan father Analysis of Denisovan DNA segments that survive in modern human populations has revealed at least three separate episodes of interbreeding between Denisovans and modern humans, each involving a genetically distinct Denisovan population.19PubMed Central. A history of multiple Denisovan introgression events in modern humans Almost everything we know about Denisovans comes from DNA rather than from fossils themselves, since the entire physical fossil record of the group consists of a finger bone, a jawbone, some teeth, and skull fragments.
Beyond human evolution, ancient DNA has rewritten the family trees of horses, dogs, cattle, and numerous wild species, revealing domestication events, migration routes, and population collapses that left no trace in the fossil record alone. It has also allowed researchers to reconstruct not just the DNA sequence but the epigenome of extinct individuals. By analyzing the damage patterns in ancient DNA, scientists have inferred which genes were switched on or off in Denisovans and Neanderthals during life, then used those methylation maps to predict anatomical features like wider fingertips and broader faces.20Cell. Reconstructing Denisovan Anatomy Using DNA Methylation Maps
When DNA Is Gone, Proteins Stick Around
DNA has a hard upper age limit. Proteins do not share that limit, at least not to the same degree. Proteins are more chemically stable than DNA and can survive over millions of years in fossils, routinely outlasting the oldest recoverable genetic material.21PubMed Central. Paleoproteomics A protein carries less information per molecule than a DNA strand, roughly one-sixth the number of atoms for the same amount of sequence data. But for fossils too old or too poorly preserved to yield DNA, protein sequences can still place an organism on a family tree. Collagen sequences extracted from bones millions of years old have been used to sort out relationships among extinct species that no amount of morphological study could resolve. Paleoproteomics is increasingly filling the gap between the DNA frontier and the deep fossil record where no genetic material survives.
De-Extinction and Its Genomic Limits
The dream of using fossil DNA to bring back extinct species runs headlong into the realities of degradation. Even at very high sequencing depths, ancient genomes have persistent gaps. A recent study modeling this problem found that roughly 3 to 4 percent of an extinct genome remains uncovered even after sequencing to 100-fold depth, and those missing regions contain functionally important variants.22PubMed Central. Mapping the Genomic Limits of De-Extinction in the Face of Ancient DNA Degradation You cannot clone an organism from an incomplete blueprint, and the missing pieces are not random; they tend to cluster in regions that are hardest to sequence, which often overlap with regions that matter biologically.
Current de-extinction strategies therefore do not try to reassemble an extinct genome from scratch. Instead, they start with a close living relative and use gene editing to introduce key traits inferred from the ancient DNA. The result is not a resurrection but an ecological proxy, an animal engineered to fill a similar role in the ecosystem.23Functional Ecology. Pathways to de‐extinction: how close can we get to resurrection of an extinct species? Even if you had every last base pair, an organism’s phenotype emerges from the interaction between its genome and the environment it develops in. A mammoth embryo gestated in an Asian elephant and raised in a modern landscape would not be the same animal as a Pleistocene mammoth. The conservation value of de-extinction may ultimately lie less in bringing back specific species and more in developing reproductive technologies, like stem cell culture and assisted reproduction, that can help endangered species alive today.24Journal of Reproduction and Development. De-extinction: how reviving the past is revolutionizing the future of conservation biology
The Cost of Sampling
Every ancient DNA study destroys part of a fossil. The standard approach involves drilling or cutting into a bone or tooth and grinding the material to powder. For teeth, this often means losing an entire root, which eliminates morphological features that other researchers might need for anatomical study, radiocarbon dating, or isotope analysis.25PubMed Central. Not a limitless resource: ethics and guidelines for destructive sampling of archaeofaunal remains Bones can break in unintended ways during sampling even when the protocol is designed to be minimally invasive.
The tension between genetic and morphological research has pushed labs to develop gentler extraction methods. One recent protocol extracts DNA from the cementum layer on the outside of tooth roots without drilling or grinding, leaving the tooth intact and suitable for further morphological and chemical study afterward.26PubMed Central. A minimally destructive protocol for DNA extraction from ancient teeth Methods like this are increasingly expected, especially for museum specimens and culturally sensitive human remains. As the ancient DNA field has grown from a handful of labs to a global enterprise, the pressure on finite fossil collections has become a real concern. A specimen can only be drilled into so many times before there is nothing left to study.