How Carbon-14 Dating Works — And Where It Falls Short

Carbon dating is probably the most publicly known application of nuclear physics. Willard Libby developed it in 1949 and won the Nobel Prize for it in 1960. The basic idea is elegant: use the steady decay of a radioactive carbon isotope as a natural clock to determine when an organism died. But the details — and the limitations — are more interesting than most people realize.

Where Carbon-14 comes from

Carbon-14 is continuously produced in the upper atmosphere. Cosmic rays collide with nitrogen atoms, knocking out a proton and replacing it with a neutron: ¹⁴N + n → ¹⁴C + p. This C-14 quickly oxidizes to CO₂ and mixes into the atmosphere, ocean, and biosphere. All living organisms absorb it through photosynthesis (plants) or through eating plants (animals). As long as an organism is alive, the ratio of C-14 to C-12 in its body stays roughly in equilibrium with the atmosphere — about 1 atom of C-14 for every 10¹² atoms of C-12.

The moment an organism dies, it stops absorbing new carbon. The C-14 it contains begins to decay via beta emission with a half-life of 5,730 years, while the stable C-12 remains unchanged. The ratio of C-14 to C-12 drops over time, and measuring that ratio tells you how long ago the organism died.

The math

Or equivalently, if you know the remaining percentage of C-14: age = −8267 × ln(fraction remaining). At 50% remaining, the age is 5,730 years (one half-life). At 25%, it's 11,460 years. At 1%, it's about 38,000 years. Our carbon dating calculator does this computation instantly.

The practical limit

After about 50,000 years (~9 half-lives), less than 0.2% of the original C-14 remains. At that point, the signal becomes indistinguishable from background radiation, and the measurement loses reliability. This means carbon dating cannot be used for dinosaur fossils (65+ million years old), geological formations, or anything older than roughly 50,000 years. For older samples, geologists use other isotopes like Uranium-238, Potassium-40, or Rubidium-87.

Why raw dates need calibration

The calculation above assumes that atmospheric C-14 levels have been constant throughout history. They haven't. Solar activity, Earth's magnetic field strength, volcanic eruptions, and ocean circulation patterns all affect how much C-14 is produced and where it ends up. During periods of low solar activity, more cosmic rays reach the atmosphere, producing more C-14. During magnetic field reversals, the same thing happens.

To correct for this, radiocarbon dates are calibrated against independent records. The primary calibration tool is tree ring data (dendrochronology) — trees produce one ring per year, and each ring's C-14 content can be measured. This record extends back about 14,000 years using bristlecone pines and European oaks. Beyond that, calibration relies on corals, lake sediments, and cave formations. The current calibration curve (IntCal20, published in 2020) extends back 55,000 years.

The difference between raw and calibrated dates can be significant — up to several hundred years in some periods. Some time intervals produce ambiguous results because the calibration curve is flat or wiggly, giving multiple possible calendar dates for a single radiocarbon measurement. The period around 400-800 AD is notoriously difficult to date precisely because of a plateau in the calibration curve.

Modern complications

Two human activities have significantly altered atmospheric C-14 levels. The burning of fossil fuels since the Industrial Revolution has diluted the atmosphere with "dead" carbon (fossil fuels are millions of years old and contain zero C-14). This is called the Suess effect, and it makes modern samples appear artificially old. Conversely, nuclear weapons testing in the 1950s and 1960s nearly doubled atmospheric C-14 levels, creating a spike called the "bomb curve." This spike is actually useful — it provides a precise time marker for samples from the second half of the 20th century and has been used in forensic science and even to study cell regeneration in the human body.

What can and can't be dated

Carbon dating works on organic material — anything that was once part of a living organism. Wood, charcoal, bone, shell, peat, seeds, textile fibers, and paper can all be dated. It does not work on rocks, metals, pottery (though organic residues on pottery can sometimes be dated), or anything that never contained biogenic carbon.

Contamination is the biggest practical challenge. A 40,000-year-old sample needs only a tiny amount of modern carbon contamination to produce a wildly wrong result. Just 1% modern contamination in a 40,000-year-old sample would make it appear 7,000 years younger. Labs spend enormous effort on sample preparation — using chemical treatments to remove contaminants before measurement. The development of Accelerator Mass Spectrometry (AMS) in the 1970s was a game-changer: it can measure C-14 directly by counting atoms rather than waiting for them to decay, requiring only milligrams of material instead of grams.

For understanding the half-life formula behind this calculation, see our step-by-step guide. For a table of all commonly used dating isotopes and their half-lives, see our isotope reference table.