Radiocarbon Dating: How Carbon-14 Tells Us the Age of Ancient Things

Photo: Roberto Lee Cortes / Pexels
In 1949, chemist Willard Libby published a paper that changed archaeology forever. He had discovered a way to read the age of ancient organic material directly from the atoms it contained — no guesswork, no stratigraphy, just physics. The technique was radiocarbon dating, and it rested on a single elegant fact: every living organism contains a tiny amount of radioactive carbon, and the moment it dies, that carbon starts a predictable countdown.
The shroud of Turin, Egyptian mummies, prehistoric cave art, Viking settlements — all of them have been interrogated by carbon-14. Understanding how the method works reveals something profound about atoms, time, and how science turns invisible decay into a historical record.
What Is Carbon-14?
Carbon exists in several forms called isotopes. All carbon atoms have six protons, but the number of neutrons varies. Most carbon on Earth is carbon-12 (six neutrons) or carbon-13 (seven neutrons). Both are stable — they do not decay.
Carbon-14 has eight neutrons, making it unstable. It decays by emitting a beta particle (an electron), converting one neutron into a proton and turning into nitrogen-14. This is radioactive decay, and it happens at a statistically predictable rate.
The half-life of carbon-14 is approximately 5,730 years. This means that in 5,730 years, half of any given amount of carbon-14 will have decayed into nitrogen-14. In another 5,730 years, half of what remains decays, and so on.
Where Does Carbon-14 Come From?
Carbon-14 is not primordial — it is constantly being created in the upper atmosphere. When cosmic rays (high-energy particles from space) collide with nitrogen-14 in the atmosphere, they knock out a proton and add a neutron, producing carbon-14. This new carbon-14 quickly oxidizes to form carbon dioxide (CO₂).
Plants absorb this radioactive CO₂ through photosynthesis, incorporating carbon-14 into their tissues at the same ratio it exists in the atmosphere. Animals eat plants (or eat animals that ate plants), so carbon-14 spreads through the entire food web. Every living organism — a tree, a woolly mammoth, a human — contains carbon-14 at the same ratio relative to stable carbon as the atmosphere does.
The key insight: as long as an organism is alive, it keeps exchanging carbon with the environment, so its carbon-14 levels stay in equilibrium with the atmosphere. The moment it dies, that exchange stops. The carbon-14 already present begins to decay, and no new carbon-14 enters.
The Clock Starts at Death
From the instant an organism dies, its carbon-14 content decreases at the known half-life rate. By measuring the ratio of carbon-14 to carbon-12 in a sample and comparing it to the ratio in the modern atmosphere, scientists can calculate how long ago the organism died.
The mathematics is straightforward:
- If a sample has 50% of the carbon-14 expected in a living organism → 1 half-life has passed → approximately 5,730 years old
- If it has 25% → 2 half-lives → approximately 11,460 years old
- If it has 12.5% → 3 half-lives → approximately 17,190 years old
The practical upper limit of radiocarbon dating is around 50,000 years. Beyond that, the remaining carbon-14 becomes too sparse to measure accurately even with modern instruments.
How Scientists Actually Measure It
Early radiocarbon dating counted individual beta particles emitted by decaying carbon-14 atoms. This required large samples — grams of material — and was slow and imprecise.
Modern dating uses Accelerator Mass Spectrometry (AMS). Instead of waiting for atoms to decay and counting the emissions, AMS physically counts the carbon-14 atoms directly by separating them from other carbon atoms using magnetic and electric fields. AMS requires only milligrams of material (a tiny fiber from a fabric, a few wood cells, a fragment of bone), and gives results with error margins as small as ±20–30 years for well-preserved recent samples.
Calibration: Correcting for the Imperfect Atmosphere
There is a complication. Carbon-14 levels in the atmosphere have not been perfectly constant over time. Solar activity variations, changes in Earth's magnetic field, and industrial fossil fuel burning (which dilutes atmospheric carbon-14 with old, carbon-14-free carbon — the Suess effect) all shift the baseline.
Scientists correct for this using calibration curves — most importantly, the IntCal curve, which was built by dating tree rings of known age (dendrochronology provides independent year-by-year age data going back over 14,000 years) and comparing the radiocarbon ages against known real ages. Coral skeletons, cave deposits (speleothems), and lake sediment varves extend the calibration further back.
The result is that radiocarbon dating does not give a simple calendar year — it gives a calibrated date range with a confidence interval, such as "1210–1390 CE at 95% confidence." For the Shroud of Turin, three independent laboratories in 1988 all returned results in that range, consistent with medieval manufacture rather than the claimed first century CE.
What Can and Cannot Be Dated
Radiocarbon dating works on any material that was once part of a living organism and has been preserved well enough:
What it dates well:
- Wood, charcoal, and plant material
- Bone and antler (using the collagen component)
- Shells and coral (though marine reservoir effects require correction)
- Seeds, textiles made from plant fibers, leather
- Peat and organic sediments
What it cannot date:
- Rocks and minerals (they were never alive and never incorporated atmospheric carbon)
- Metals and ceramics (though associated organic material like charcoal from a kiln can be dated)
- Material older than ~50,000 years (too little carbon-14 remains)
- Living organisms (the ratio is still in equilibrium — no age signal)
For dating older geological materials, scientists use other radiometric methods with longer half-lives: potassium-argon dating (half-life ~1.25 billion years) for volcanic rock, uranium-lead dating for very ancient minerals, and rubidium-strontium dating for rocks billions of years old.
Famous Applications
Ötzi the Iceman — The remarkably preserved Neolithic man found in the Alps in 1991 was dated to approximately 3300 BCE, placing him in the Copper Age and making him Europe's oldest known natural human mummy.
Dead Sea Scrolls — Carbon-14 dating of linen wrappings confirmed the scrolls' origin between roughly 250 BCE and 50 CE, consistent with the historical period implied by their content. This helped authenticate the most significant biblical manuscript find of the 20th century.
Lascaux Cave Paintings — The cave art at Lascaux in France was dated using charcoal found in the paintings themselves to approximately 17,000 years ago, placing its creation firmly in the Upper Paleolithic.
Kennewick Man — A controversial ancient skeleton found in Washington State in 1996 was dated to approximately 8,500 years old, helping resolve debates about early human migration into the Americas.
The marine reservoir effect and atmospheric corrections
Radiocarbon dating assumes all organisms exchange carbon with the atmosphere at the same rate. But marine organisms — fish, shellfish, seaweed — draw carbon from dissolved inorganic carbon in the ocean rather than the atmosphere. Ocean water circulates slowly, and deep water contains "old" carbon that has been isolated from the atmosphere for centuries. This means marine organisms are typically older by 300–400 years when dated by radiocarbon than their actual age — the "reservoir age."
Scientists correct for this using calibration curves specific to marine samples, but the correction is imperfect when the exact water origin is unknown. A shell collected from the coast today contains carbon from both surface waters (recently exchanged with atmosphere) and deeper waters (older), creating residual uncertainty even after correction.
This effect is particularly important in archaeological sites near coasts or based on shellfish middens. Japanese archaeologists dating Jomon shell middens must apply marine reservoir corrections to avoid systematic errors of centuries. Conversely, freshwater reservoirs (lakes and rivers) may have intermediate ages depending on circulation and carbon input from surrounding land — requiring specific correction factors.
The Bayesian approach to calibration, implemented in software like OxCal and Calib, can incorporate multiple dates from a single site or stratigraphic sequence and resolve ambiguities. If ten radiocarbon dates are taken from successive layers in a sediment core, the Bayesian model uses the physical constraint that deeper layers must be older, and can reconcile apparent inconsistencies by identifying contaminated samples or unaccounted reservoir effects.
Source: Radiocarbon Journal (2023) — Marine Reservoir Effects and Bayesian Calibration Methods.
Beyond radiocarbon: Other radiometric dating methods
Radiocarbon is powerful but limited to ~50,000 years. For older materials, scientists use isotope systems with longer half-lives.
Potassium-argon (K-Ar) dating — Potassium-40 decays to argon-40 with a half-life of 1.25 billion years. This method dates volcanic minerals and is used for dating lava flows that bracket fossils (like hominins in the Turkana Basin of Kenya, dated via K-Ar analysis of volcanic layers). The method requires extremely careful measurement — even trace atmospheric argon contamination can introduce errors.
Uranium-lead (U-Pb) dating — Uranium-238 decays to lead-206 over 4.5 billion years; uranium-235 to lead-207 over 700 million years. U-Pb dating is applied to zircons (crystals containing uranium as an impurity) and is among the most accurate methods for dating ancient rocks and meteorites. The Hadean zircons from Jack Hills, Australia (dated to 4.4 billion years) provide evidence that Earth's crust formed nearly as soon as the planet did.
Rubidium-strontium (Rb-Sr) dating — Rubidium-87 decays to strontium-87 with a half-life of 49 billion years. Used for dating rocks billions of years old. The technique revealed that the Earth and meteorites are all approximately 4.54 billion years old — the age of the solar system itself.
Samarium-neodymium dating — Used for dating meteorites and the oldest terrestrial rocks, with even longer half-lives than Rb-Sr.
Each method has advantages and limitations. K-Ar works on young volcanic rocks; U-Pb on ancient minerals. All require careful sample preparation, laboratory standards, and chemical analysis.
Source: USGS — Radiometric Dating: A Practitioners' Guide to Radiometric Dating Methods.
The calibration curve: Reading 14,000 years of history
The calibration curve is the backbone of accurate radiocarbon dating. It maps radiocarbon ages against calendar ages, accounting for past variations in atmospheric carbon-14 concentration. Building the curve required exquisite detective work.
Dendrochronology (tree-ring dating) provided the anchor. Each year, a tree adds a ring; the width varies with climate. By overlapping ring patterns between living trees and dead logs, researchers extended the dendrochronological record to over 14,000 years ago (the bristlecone pines of California and Mediterranean oak sequences). Samples of known calendar age (from dendrochronology) were radiocarbon-dated, revealing that atmospheric carbon-14 concentration has varied by ±5–15% over the past 14,000 years.
Beyond 14,000 years ago, marine corals and lake sediment varves (annual layers) provided calibration data out to ~55,000 years. Speleothems (cave formations like stalactites), which can be dated both by radiocarbon (using dissolved organic carbon) and uranium-thorium dating (an independent isotope system), provided additional cross-checks.
The result is the IntCal curve (International Calibration) and regional variants (SHCal for Southern Hemisphere, MarineCal for marine samples). Every few years, the curve is updated as more high-precision data accumulates. IntCal20 (released 2020) incorporates over 700 radiocarbon dates with independently known ages.
A radiocarbon result of "3500 ± 50 BP" (before present, i.e. before 1950 CE) is statistically calibrated against IntCal20, yielding a calendar age range like "3760–3650 BCE at 68% confidence." This range accounts for both the radiocarbon measurement uncertainty and the fact that the calibration curve has "wiggles" — flat sections and reversals where atmospheric carbon-14 fluctuated.
Challenges and contamination
Despite radiocarbon dating's power, several artifacts can skew results.
Fractionation — different chemical forms of carbon (¹²C vs ¹³C vs ¹⁴C) are incorporated by organisms at slightly different rates. Living wood preferentially incorporates lighter isotopes (kinetic isotope effect). Laboratories measure the ratio of ¹³C to ¹²C in a sample and apply small corrections. Ignoring fractionation can introduce errors of 50–100 years.
Contamination — the most common cause of unreliable dates. A 10,000-year-old sample contaminated with 1% modern carbon appears 500 years younger than its true age. Contaminants include rootlets (modern plant roots penetrating archaeological layers), dust, fungal growth, and humic acids migrating from upper soil layers. Modern labs use AMS with chemical pretreatment: acid/base/acid washes to remove carbonate and humic contaminants, then combustion to pure CO₂ before AMS analysis.
Hard water effects — in regions with limestone, groundwater contains carbonate from dissolved rock, which is carbon-14-depleted (infinitely old). Shells and bones exposed to hard water may incorporate old carbon, appearing older than they are.
Old-wood problem — a wooden beam in an ancient building may be centuries older than the structure itself, if harvested from a long-lived tree. A beam dated to 100 CE could have been used in a structure built 200 years later. Contextual archaeology (dating seeds or textiles found inside, not the wood itself) avoids this pitfall.
Experienced radiocarbon labs anticipate these issues and report comprehensive uncertainty budgets, not just the radioactive decay uncertainty. Multiple samples from independent contexts are always preferable to relying on a single measurement.
Source: Radiocarbon (2024) — Best Practices in Radiocarbon Sample Preparation and Interpretation.
Archaeological chronologies: Bayesian dating and stratigraphy
Individual radiocarbon dates are useful, but their true power emerges when combined with archaeological context. Bayesian chronological modelling allows researchers to simultaneously process dozens or hundreds of radiocarbon measurements from a single site, using physical and stratigraphic constraints to resolve ambiguities and refine dates.
Consider an archaeological site with multiple occupation layers. Layer A is physically above layer B (stratigraphy dictates layer B is older), yet five radiocarbon measurements from layer B yield a spread of dates, and one measurement from layer A appears older than some layer B dates. The naive approach — treating each date as independent — would accept this contradiction. The Bayesian approach uses software like OxCal or JAGS (Just Another Gibbs Sampler) to enforce stratigraphic ordering. It solves for the most likely occupation period for each layer, identifying which measurement is an outlier (perhaps contaminated) and resolving the apparent contradiction while accounting for calibration curve wiggles.
This technique revolutionised understanding of human migration timelines. The expansion of Polynesian settlement across the Pacific, the peopling of the Americas, and the Neolithic transition in Europe were all clarified by Bayesian radiocarbon analysis. For instance, dating the Valdivia culture in Ecuador (one of the world's earliest pottery traditions) relied on precise Bayesian refinement of radiocarbon dates from multiple house floors and midden layers, constraining Valdivia's emergence to 3200–3000 BCE with narrow confidence intervals.
Radiocarbon wiggle-matching is another technique: if you have a series of annual tree rings (from dendrochronology or precision measurement of wood samples), you can radiocarbon-date multiple rings and compare the resulting radiocarbon curve to known wiggles in the IntCal calibration curve. When the patterns match, you've pinpointed the calendar age of those rings to within a few decades, even if they're not attached to a living-tree sequence. This has been essential for dating Neolithic and Bronze Age sites in regions without continuous dendrochronological records. Combined with other radiometric dating methods, radiocarbon provides the temporal framework for understanding archaeological and paleontological sequences across millennia.
Source: Bronk Ramsey, C. (2009) Bayesian analysis of radiocarbon dates; Oxford Journal of Archaeology special issue on chronology methods (2018).
Libby's legacy and the future of radiocarbon dating
Willard Libby won the Nobel Prize in Chemistry in 1960 for developing radiocarbon dating. His insight was to use the predictability of radioactive decay — a purely quantum mechanical phenomenon — as a timekeeping device. What might seem a limitation (the random, probabilistic nature of radioactive decay) becomes an advantage at scale: with billions of atoms in even a tiny sample, the statistical average is extremely predictable.
Radiocarbon dating is a beautiful illustration of how processes happening in the upper atmosphere (cosmic ray bombardment), global biological cycles (photosynthesis and food web), and nuclear physics (radioactive decay) are interconnected. Reading the age of a 3,000-year-old piece of wood means understanding all three.
Future advances include compound-specific radiocarbon dating — measuring carbon-14 in individual amino acids or fatty acids within a sample, allowing dating even when bulk samples are contaminated. Ultra-high-resolution AMS now requires only micrograms of material (billionths of a gram for single compounds), opening new possibilities for dating rare or precious artifacts without visible damage.
Radiocarbon dating remains one of science's most powerful tools for bridging the gap between the geological deep past (dated by uranium-lead in rocks billions of years old) and recorded history (documented in texts). For archaeology, paleontology, climate science, and geology, it has been indispensable for nearly 75 years.
Frequently Asked Questions
Q: How accurate is radiocarbon dating?
A: For samples younger than 10,000 years and properly calibrated with pretreatment to remove contaminants, radiocarbon dating can be accurate to within decades — sometimes ±50 years or better. For older samples, uncertainty grows: a 30,000-year-old sample might have uncertainty of several hundred years, which is still remarkably precise given the timescale. Accuracy depends on contamination control and calibration curve quality.
Q: Can radiocarbon dating be deliberately fooled or faked?
A: A sample can be naturally contaminated with older or newer carbon, which shifts apparent age. Modern laboratories use rigorous pretreatment protocols (acid/base/acid washes, chemical separation) to remove contaminants. Deliberate faking would require replacing the carbon atoms throughout an artifact while preserving all other features — essentially impossible without completely destroying and reconstructing the object.
Q: Why can't radiocarbon dating go back further than 50,000 years?
A: After about 8–9 half-lives (~50,000 years), less than 0.1% of the original carbon-14 remains. The residual signal becomes indistinguishable from background radiation, cosmic rays hitting AMS equipment, and measurement noise. Older materials require isotope systems with much longer half-lives: potassium-argon (1.25 billion year half-life) or uranium-lead (billions of years).
Q: Does radiocarbon dating work on dinosaur bones?
A: No. Dinosaurs went extinct approximately 66 million years ago — roughly 1,300 times longer than radiocarbon's 50,000-year limit. Any detectable carbon-14 in dinosaur samples comes from contamination (modern carbon from handling, soil, or lab environments). Researchers claiming radiocarbon finds in dinosaurs are measuring background contamination, not original material.
Q: How did scientists build the calibration curve going back 14,000 years?
A: Dendrochronology (tree-ring dating) provided the primary anchor. By overlapping ring patterns from living trees, dead logs, and ancient dead wood, researchers extended the record to ~14,600 years. Beyond that, marine corals (dated independently with uranium-thorium dating) and lake sediment varves (annual layers) extended calibration to ~55,000 years. The result is the IntCal curve, updated every few years as new high-precision data is added.
Q: Can radiocarbon dating work on extremely small samples?
A: Yes — modern accelerator mass spectrometry (AMS) requires only milligrams or even micrograms of material. Compound-specific radiocarbon dating can date individual amino acids or lipids within samples, useful for rare or precious artifacts. Single-compound dating improves accuracy by isolating specific molecules away from contamination, though it requires sophisticated chemistry and specialised equipment only available at a few dozen laboratories worldwide.
Related Reading
Follow new articles
Add SciHub101 to your feed reader to see new science and technology stories.
About the Author
SciHub101 is an independent science and technology publication founded and edited by Buğra Sözeri, a software engineer and app developer based in Türkiye. AI tools assist with drafting and research. Source documentation in the existing archive is uneven, and articles are being reviewed and corrected. The Editorial Team byline identifies editorial responsibility; it does not certify specialist or independent human review. Where we compare products, anything built by our own team is labelled as such — see the editorial guidelines and About page for how corrections and disclosures work.
Related Articles

The Feynman Technique — How to Learn Anything
Richard Feynman's 4-step method for understanding hard ideas. Why it works, what cognitive science says, and how it compares to other study techniques.

Plate Tectonics Explained — How and Why Earth's Crust Moves
Plate tectonics explained: how Earth's rigid plates move, what drives them, what happens at their boundaries, and the evidence that settled the debate.

Complete Science Guide 2026: Physics to Climate
How physics, chemistry, earth and climate science and materials science work, with the evidence behind each field and the best deep dive for every topic.

How Radioactive Decay Works: Half-Life, Carbon Dating, and Nuclear Medicine
Atoms are not all stable. Some nuclei spontaneously transform, emitting radiation as they go. The process powers nuclear medicine, reveals the age of ancient artifacts, and tells us the age of the Earth itself.

Marie Curie: Discovering Radioactivity and Redefining Science
Marie Curie's discovery of radioactivity and isolation of radium revolutionized physics and chemistry. The first woman to win a Nobel Prize, she sacrificed health for science.

What If We Ran Out of Oxygen for 5 Seconds? — Science Explained
What would actually happen if Earth's oxygen vanished for just 5 seconds? Buildings would implode, the sky would go black, your eardrums would rupture, and all metal surfaces would weld together. Oxygen is doing far more than you think.