# How long does DNA last?

> DNA can last months on an outdoor surface, centuries in buried bone, and about 2 million years in frozen sediment. Temperature, water, sunlight, and microbes set the pace far more than age does.

A bloodstain left outdoors can stop yielding a complete DNA profile within a few months. A mammoth tooth frozen in Siberian permafrost still held readable DNA after more than a million years. Both are fair answers to how long DNA lasts, because DNA has no fixed lifespan. It breaks into ever shorter, chemically damaged fragments at a pace set by temperature, water, sunlight, microbes, and the material around it.

The figure most often quoted is a **521-year half-life**, measured in New Zealand moa bones for one stretch of mitochondrial DNA 242 base pairs long. A base pair is one rung of the twisted DNA ladder. That half-life is a decay rate, not an expiry date. The oldest DNA sequenced so far, from frozen sediment in northern Greenland, is about 2 million years old.

## How long does DNA last?

DNA lasts anywhere from a few months on an outdoor surface to about 2 million years in frozen sediment. Table 1 collects the best-measured cases, ordered roughly from the shortest survival to the longest.

  <caption><strong>Table 1. How long DNA lasts in different settings.</strong> A complete profile is a full forensic DNA match pattern for the person who left the sample. Values describe recoverable fragments, not intact molecules. Sources: Poetsch et al. (2022), Arsenault et al. (2025), Ivanova and Kuzmina (2013), Spalding et al. (2005), Allentoft et al. (2012), van der Valk et al. (2021), Kjær et al. (2022).</caption>
  <thead>
    <tr>
      <th scope="col">Setting</th>
      <th scope="col">How long the DNA lasted</th>
      <th scope="col">Conditions</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Blood or saliva on a surface outdoors</td>
      <td>Under 25% complete profiles after 3 months, none after 12 months</td>
      <td>Cloth and plastic exposed to sun, rain, and seasonal temperatures</td>
    </tr>
    <tr>
      <td>Blood, saliva, or skin cells indoors</td>
      <td>Nearly all blood and saliva samples complete for 9 months; skin-cell profiles declined after 3 months</td>
      <td>Cloth and plastic kept in rooms, summer and winter</td>
    </tr>
    <tr>
      <td>Trace DNA on household materials</td>
      <td>Some DNA recovered after 1 year on seven of eight surfaces; none on wood after 4 months</td>
      <td>Known deposits on non-metal surfaces under three storage conditions</td>
    </tr>
    <tr>
      <td>Purified DNA stored dry</td>
      <td>Still amplified and sequenced after 4 years</td>
      <td>Room temperature; frozen controls gave longer reads</td>
    </tr>
    <tr>
      <td>DNA in a living brain neuron</td>
      <td>The person's whole life</td>
      <td>Kept by continual repair inside the cell</td>
    </tr>
    <tr>
      <td>Mitochondrial DNA in buried bone</td>
      <td>Half-life of 521 years for a 242-base-pair fragment</td>
      <td>Moa bones at an estimated 13.1°C</td>
    </tr>
    <tr>
      <td>Short fragments in frozen bone</td>
      <td>Predicted half-life of 158,000 years for 30 base pairs</td>
      <td>Model prediction for bone at -5°C</td>
    </tr>
    <tr>
      <td>Mammoth teeth in permafrost</td>
      <td>Genome-wide data from specimens more than 1 million years old</td>
      <td>Permafrost in eastern Siberia</td>
    </tr>
    <tr>
      <td>Frozen sediment in Greenland</td>
      <td>About 2 million years</td>
      <td>Mean annual temperature of -17°C; DNA bound to clay minerals</td>
    </tr>
  </tbody>

Two things explain the spread. The first is the setting: each step down the table is colder, drier, darker, or better shielded than the one above it. The second is what counts as lasting. A forensic laboratory needs enough DNA, in long enough pieces, to read many places in the genome at once, so a stain can fail long before its last fragments disappear.

Ancient-DNA researchers ask much less of a sample. They can work with fragments a few dozen base pairs long and piece information together from millions of them. The Greenland record rests on DNA modeled to average about 50 base pairs. The rest of this article follows DNA from the living cell to the fossil and then to the crime scene and the laboratory freezer.

## How long does DNA last in a living cell?

Inside a living cell, DNA lasts as long as the cell does, because the cell repairs damage as fast as it appears. The molecule is not chemically stable on its own. Lindahl showed that hydrolysis by water, oxidation, and nonenzymatic methylation damage DNA at significant rates inside cells, and that specific repair processes counteract them.

A cell therefore keeps its original DNA instead of replacing it, unlike [messenger RNA](/guides/how-long-does-mrna-last-in-a-cell), which is broken down and remade all the time. Spalding and colleagues measured how old that DNA can be. Atmospheric nuclear weapons tests raised carbon-14 levels sharply until 1963, and the levels have fallen ever since. Because the carbon-14 in genomic DNA tracks the atmosphere at the time the DNA was made, it dates when a cell was born.

Neurons in the occipital cortex, at the back of the brain, carried DNA as old as the person. That DNA had been maintained for an entire lifetime, while the non-neuronal cells around it had been replaced. Maintenance stops at death, and from that moment the clock described in the rest of this article starts.

## How long does DNA last after death?

DNA starts to degrade soon after death, but recoverable fragments can persist from days to more than a million years. The rate depends more on the tissue and its surroundings than on elapsed time alone.

Once repair stops, the cell's own nucleases cut DNA, microbes digest the tissue, and water-driven reactions remove bases and break the sugar-phosphate backbone. Ancient DNA is therefore rarely intact. A review of specimens between 4 and 13,000 years old found that almost all recovered DNA had been reduced to fragments 40 to 500 base pairs long, with chemical damage concentrated near the fragment ends.

Detectable DNA is not the same as an intact genome. Each human cell holds about [2 meters of DNA](/guides/how-long-is-dna) when stretched out, and a [human genome contains about 3.1 billion base pairs](/guides/human-genome-size). Sequencing can still read many short fragments and place them against a reference sequence, recovering information long after the long molecules have broken apart.

Dense tissues shield more DNA than soft tissue. One comparison of 34 ancient skeletons found average endogenous DNA contents, the share of DNA from the person rather than from soil microbes, of 40.0% in inner petrous bone, 16.4% in tooth cementum, and 2.2% in parietal skull bone. Preservation still varied widely among individuals and burial sites.

## What is the half-life of DNA?

The best-measured half-life of DNA is 521 years, for a 242-base-pair fragment of mitochondrial DNA in bone buried at about 13°C. It comes from 158 radiocarbon-dated moa bones from New Zealand, aged 602 to 7,839 years and collected from three deposits less than 5 kilometers apart.

A half-life means that the number of intact copies of that fragment fell by half every 521 years in the fitted model. After 1,042 years, one-quarter would remain; it does not mean that all DNA disappears after 521 years. The measured rate was almost 400 times slower than predicted from laboratory measurements of depurination, the loss of DNA bases, a sign of how different conditions in bone are from those in a test tube.

The fragment studied was [mitochondrial DNA](/guides/how-many-genes-in-mitochondrial-dna), a small circular genome present in many copies per cell. Nuclear DNA decayed at least twice as fast in the two specimens tested by high-throughput sequencing, although that comparison rests on only two samples. A single half-life cannot describe every genome, tissue, or environment.

The same model predicts far longer survival for shorter fragments in colder bone. At -5°C, a 30-base-pair fragment would have a half-life of 158,000 years.

![Predicted half-life of a 30-base-pair mitochondrial DNA fragment in bone at temperatures from -5°C to 25°C](/images/charts/dna-fragment-half-life-by-temperature.webp '**Figure 1. Predicted half-life of a 30-base-pair DNA fragment in bone.** Model predictions from the moa bone decay rate, on a logarithmic scale. Source: [Allentoft et al. (2012)](https://pmc.ncbi.nlm.nih.gov/articles/PMC3497090/).')

From these predictions, we calculate that each 10°C of cooling lengthens the half-life six- to eightfold, from 500 years at 25°C to 158,000 years at -5°C. Even at -5°C, the authors predicted that no intact bonds would remain after 6.8 million years. They treated this as a limit under the best conditions and suggested that readable fragments may survive more than 1 million years in deep frozen environments.

## What conditions preserve DNA the longest?

Cold, dry, dark, chemically stable conditions preserve DNA best. Heat, moisture, ultraviolet light, oxygen, acidity, and microbial activity speed up its loss.

Temperature has an especially large effect because it controls the rates of chemical reactions and microbial growth. Water enables hydrolysis, while sunlight causes direct photochemical damage. Burial chemistry matters too: pH, salts, oxygen, and contact with minerals can change how quickly DNA breaks down.

Until 2022, the oldest DNA known came from mammoth teeth preserved in Siberian permafrost. Van der Valk and colleagues recovered genome-wide data from three mammoth specimens, two of them more than a million years old.

The current record is environmental DNA, shed by plants and animals into marine sediment at Kap København in northern Greenland about 2 million years ago. The researchers identified poplar, birch, mastodons, and reindeer by matching each short fragment against databases of known sequences, the same comparison a [BLAST search](/app/blast-search) makes for a single sequence.

Cold made the record possible. Using the site's mean annual temperature of -17°C, the team calculated that 2 million years of exposure was equivalent to only about 2,700 years at a constant 10°C. Clay minerals, particularly smectite, also bound DNA far more strongly than quartz, and adsorbed DNA is less accessible to the enzymes that degrade it.

The 2-million-year record does not mean that DNA normally lasts that long. A young sample kept warm and wet may yield less usable DNA than a much older one held cold and dry.

## How long does DNA last on surfaces and at crime scenes?

Indoors, blood and saliva on a surface can still give a complete DNA profile after about 9 months. Outdoors, most samples fail after 3 months, and in one long-term study none gave a complete profile after a year.

Poetsch and colleagues placed 338 skin-cell, 572 blood, and 572 saliva samples on cloth and plastic, then left them indoors or outdoors through summer and winter. A complete profile means the laboratory could read every site in the standard forensic panel of repeated DNA stretches. Indoors, nearly all blood and saliva samples still gave one at 9 months, while skin-cell samples began to decline after 3.

Outdoor exposure was far harsher. After 3 months, fewer than 25% of blood and saliva samples gave a complete profile, and after 12 months none did. The study tracked sunlight, temperature, and humidity, all of which vary far more outdoors than in a room.

The surface matters as much as the weather. A 2025 controlled study placed cellular and cell-free DNA on eight non-metal surfaces under three storage conditions. Some DNA remained recoverable for the full year on every surface except wood, which yielded none after four months. Cellular DNA persisted better on porous materials and cell-free DNA on non-porous ones.

These experiments used known deposits under defined conditions. Real traces can be much smaller, touched or cleaned, or mixed with DNA from other people, so the laboratory durations are best read as upper limits rather than typical values.

## How long does stored DNA last?

Purified DNA kept dry lasts for years at room temperature, and a freezer keeps it in better condition. In a four-year comparison, dry room-temperature DNA could still be amplified and sequenced, but samples kept at -20°C gave longer reads and stronger signals. The authors saw dry storage as a backup and a way to ship samples, not a replacement for cold archives.

A second study followed purified DNA for up to 365 days. DNA dried in a commercial protective matrix at room temperature recovered about as well as frozen controls and better than unprotected DNA. Accelerated-aging tests suggested the protection matters most for samples under 20 nanograms.

Across all of these settings, DNA is always falling apart; what changes is the speed. A living cell outpaces the damage with repair, and a freezer, a dry tube, or permafrost slows it down. Whether DNA has lasted then depends on the purpose: a full forensic profile needs far more than the fragments of about 50 base pairs that identified a 2-million-year-old forest.
