How Old Is the Universe — And How Do We Actually Know?

The universe is about 13.8 billion years old. This number appears in textbooks and science documentaries with such confidence that it's easy to miss what an extraordinary achievement it represents. How do you measure the age of something that contains everything — including all possible clocks, all possible rulers, and all possible observers?

The answer is not a single measurement. It is the convergence of multiple independent methods, using different physics, all pointing to roughly the same answer. That convergence is what makes cosmologists confident — and the places where the methods do not quite agree are where the most interesting physics is hiding.

Key takeaways

• The universe's age (about 13.8 billion years) is supported by three independent lines of evidence: the cosmic microwave background, the ages of the oldest stars, and the expansion rate (Hubble constant).
• The cosmic microwave background (CMB) gives the most precise value: Planck's 2018 baseline analysis found 13.797 ± 0.023 billion years, assuming the standard ΛCDM model.
• The oldest stars come out at roughly 12–14.5 billion years, with uncertainties of several hundred million years — consistent, but much less precise.
• The Hubble tension — a gap between the CMB-inferred expansion rate (67.4 km/s/Mpc) and the local distance-ladder value (about 73 km/s/Mpc) — is one of the biggest open problems in cosmology.
• The age defines the observable universe: light has had about 13.8 billion years to travel, but because space expanded meanwhile, the observable universe's radius today is about 46 billion light-years.

A brief history of the question

Until the twentieth century most scientists assumed the universe was eternal and unchanging. The question "how old is it?" only became scientific once astronomers found that it was changing.

Year	Development	Implied age
1912	Henrietta Leavitt publishes the period–luminosity law for Cepheid variable stars, the first reliable cosmic yardstick	—
1927	Georges Lemaître derives an expanding universe from general relativity and estimates the expansion rate	—
1929	Edwin Hubble publishes the velocity–distance relation for galaxies	~2 billion years (his distances were badly underestimated)
1952–1958	Walter Baade and Allan Sandage revise Cepheid and galaxy distances	Rose to roughly 10–20 billion years
1965	Arno Penzias and Robert Wilson detect the cosmic microwave background	Big Bang confirmed; precise age still unknown
1990s	Hubble Space Telescope Key Project narrows the expansion rate; age ~12–14 billion years	~13 billion years
2003–2013	WMAP maps the CMB in detail	13.77 billion years (9-year results)
2013–2018	Planck satellite releases its final CMB results	13.8 billion years

Hubble's original estimate of about 2 billion years was an embarrassment: geologists already knew Earth was older than that from radioactive dating. The problem was the distance scale. Hubble had confused two different populations of Cepheid stars, and correcting that — mostly the work of Baade in the 1950s — roughly doubled the size and age of the universe at a stroke. The story is told in more detail in our profiles of Edwin Hubble and Henrietta Leavitt, whose Cepheid law made the whole enterprise possible.

Method 1: The cosmic microwave background

The most precise measurement of the universe's age comes from the cosmic microwave background (CMB) — the afterglow of the Big Bang, first detected by accident in 1965, as described in the cosmic microwave background discovery.

About 380,000 years after the Big Bang, the universe cooled enough for protons and electrons to combine into neutral hydrogen atoms. Before this, the universe was an opaque plasma — photons couldn't travel freely, continually scattered by free electrons. At this "recombination" epoch, the universe became transparent, and photons were released to travel freely through space.

Those photons have been travelling ever since, cooling as the universe expanded. Today they form the CMB — a nearly perfectly uniform glow of microwave radiation coming from all directions, at a temperature of about 2.725 K (−270.4°C).

The CMB is not perfectly uniform. There are tiny temperature variations — anisotropies — of about one part in 100,000. These variations encode the state of the universe at recombination: the distribution of matter, the strength of early density fluctuations, and the fundamental cosmological parameters.

How a picture of the baby universe becomes an age

The CMB does not contain a time stamp. The age comes out indirectly, in three steps:

1. Measure the pattern. Satellites map the temperature (and polarisation) across the whole sky and compute the power spectrum — how strong the fluctuations are at each angular size. The spectrum has a series of peaks produced by sound waves sloshing through the early plasma.
2. Fit a model. The positions and heights of those peaks depend on a small set of numbers: the densities of ordinary matter, dark matter and dark energy, the expansion rate, and the properties of the initial fluctuations. Cosmologists find the values of the six base parameters of ΛCDM that best reproduce the observed spectrum.
3. Integrate the history. With those densities fixed, the Friedmann equations of general relativity say exactly how fast the universe expanded at each epoch. Adding up the time elapsed from the Big Bang to the present gives the age.

The key physical ruler is the sound horizon: the distance a sound wave could travel in the plasma before recombination, about 145 megaparsecs today. Its apparent angle on the sky is measured extraordinarily precisely, and it pins down the geometry and expansion history.

Planck satellite results (2018): the baseline analysis (temperature, polarisation and lensing) gives an age of 13.797 ± 0.023 billion years. Adding baryon acoustic oscillation data from galaxy surveys gives 13.787 ± 0.020 billion years. Either way, the statistical uncertainty is about 0.2% — remarkable for measuring something 13.8 billion years old.

The crucial caveat is that this is a model-dependent number. It is the age if ΛCDM is the right description of the universe's contents and history. That is exactly why independent methods matter.

Method 2: The oldest stars

Stars are cosmic clocks. Their structure and chemistry change in predictable ways as they burn through their fuel, and no star can be older than the universe it lives in. Star ages therefore set a lower limit on cosmic age that is entirely independent of cosmology.

The oldest known stars are in globular clusters — ancient, densely packed groups of hundreds of thousands of stars orbiting the halo of the Milky Way, described in what globular clusters are. These clusters formed in the early universe and their stars contain very little iron and other heavy elements, a sign they formed before many generations of supernovae had enriched the gas.

Main-sequence turnoff dating

Stellar age measurement in clusters uses the main-sequence turnoff point. In a Hertzsprung-Russell diagram (luminosity vs temperature), the stars of a cluster — all born at about the same time — trace a main sequence. Massive stars burn their hydrogen faster, so as a cluster ages, stars peel off the top of the main sequence toward red giants from the most massive downward. The point at which the cluster's main sequence "turns off" tells you the cluster's age: the stars turning off now are exactly old enough to have used up their core hydrogen.

The method depends on knowing the cluster's distance (to get true luminosities) and on stellar models. Gaia's precise parallaxes have improved the distances considerably; the remaining uncertainties lie in the physics of stellar interiors, such as how heavy elements settle and how convection mixes the core.

Single old stars: the "Methuselah star"

The nearby subgiant HD 140283, nicknamed the Methuselah star, became famous in 2013 when a team led by Howard Bond used Hubble Space Telescope measurements of its distance to estimate an age of 14.5 ± 0.8 billion years. Taken at face value that is older than the universe, but the uncertainty overlaps comfortably with 13.8 billion years, and a 2014 reanalysis lowered the central value to about 14.3 billion years. The star is a reminder that individual stellar ages carry uncertainties of hundreds of millions of years.

Other stellar clocks

• White dwarf cooling. White dwarfs are stellar embers that cool at a calculable rate. The faintest, coolest white dwarfs in a cluster give an independent age — see white dwarfs.
• Radioactive dating of stars (nucleocosmochronology). A few very old stars show spectral lines of thorium and uranium. Because these isotopes decay with known half-lives, comparing their abundances with stable elements yields an age, in the same way carbon dating works for organic material.

Results: Globular cluster ages from recent studies generally fall in the range of about 12 to 13.5 billion years, with uncertainties of several hundred million years. These numbers are slightly below 13.8 billion years, which is what you would expect: the first stars and clusters formed a few hundred million years after the Big Bang, not at t = 0.

Method 3: The Hubble constant and cosmic expansion

The universe is expanding. In 1929, Edwin Hubble established that galaxies are receding from us at a speed proportional to their distance:

v = H₀ × d

where H₀ is the Hubble constant — the proportionality between recession velocity and distance. The Hubble constant has units of km/s/Mpc (kilometres per second per megaparsec of distance; one megaparsec is about 3.26 million light-years).

If you know how fast the universe is expanding, you can project backward to find when everything was at the same point — the Big Bang. A faster expansion implies a younger universe (it expanded to its current size more quickly).

A worked example

The quantity 1/H₀ is called the Hubble time. Converting units:

• For H₀ = 70 km/s/Mpc, 1/H₀ ≈ 14.0 billion years.
• For H₀ = 67.4 km/s/Mpc, 1/H₀ ≈ 14.5 billion years.
• For H₀ = 73 km/s/Mpc, 1/H₀ ≈ 13.4 billion years.

The true age is not exactly the Hubble time, because the expansion rate has changed. Gravity from matter slowed the expansion for the first several billion years; dark energy has been accelerating it since. In a universe with about 31.5% matter and 68.5% dark energy (the Planck values), these effects nearly cancel, and the age is about 0.95 × 1/H₀. With H₀ = 67.4, that gives roughly 13.8 billion years. Keeping the same composition but using H₀ = 73 would give roughly 12.7–12.8 billion years.

Measuring H₀: the tension

This is where cosmology currently has an unresolved problem.

Method	Team / dataset	H₀ (km/s/Mpc)	Year
CMB + ΛCDM model	Planck Collaboration	67.4 ± 0.5	2018
Cepheids + Type Ia supernovae	SH0ES (Riess et al.), 42 supernova hosts	73.04 ± 1.04	2022
Tip of the red giant branch (TRGB) + supernovae (HST and JWST)	Chicago-Carnegie Hubble Program (Freedman et al.)	70.39 ± 1.22 (stat) ± 1.33 (sys) ± 0.70 (SN Ia)	2024–25
Gravitational-wave standard siren (GW170817)	LIGO/Virgo and partners	70.0 (+12.0 / −8.0)	2017

• CMB-based (indirect) measurement: Using Planck CMB data and the standard cosmological model, H₀ = 67.4 ± 0.5 km/s/Mpc. This is a prediction of today's expansion rate from conditions 13.8 billion years ago.
• Local distance ladder (direct) measurement: Using Cepheid variable stars to calibrate the brightness of Type Ia supernovae, then measuring recession velocities of distant galaxies. The SH0ES team's 2022 result, 73.04 ± 1.04 km/s/Mpc, differs from Planck at about the 5-sigma level.

This roughly 8% discrepancy — the Hubble tension — is one of the most significant open problems in cosmology. The two methods use different physics and data. Either there's an unknown systematic error in one or both measurements, or there's new physics not included in ΛCDM.

Measurements using other stars as distance indicators, led by Wendy Freedman's group, land in between, at around 68–70 km/s/Mpc, which some researchers read as evidence that the Cepheid ladder has a hidden systematic. The SH0ES team disputes this. As of 2026 the question is not settled.

The Hubble tension: could the universe be younger?

If H₀ is actually 73 km/s/Mpc and nothing else changed, the universe's age would be closer to 12.7–12.8 billion years — younger than the CMB measurement suggests and uncomfortably close to the ages of the oldest globular clusters.

In practice, most proposed solutions do not simply raise H₀ while keeping everything else fixed. They change the early universe so that the CMB fit itself gives a higher H₀ — and that changes the inferred age too. Possible explanations include:

• Unknown systematics in Cepheid distance calibration, crowding of stars in distant galaxies, or supernova standardisation
• Early dark energy — an extra component of energy in the first few hundred thousand years that would shrink the sound horizon
• Interacting dark energy or dark matter models — see dark matter and dark energy explained
• Extra relativistic particles in the early universe, beyond the three known neutrino species

This is why star ages matter so much. A firm lower limit from the oldest stars restricts how young the universe can be, and therefore which proposed solutions are allowed.

Until the tension is resolved, some uncertainty remains in the precise age of the universe — though 13.8 billion years remains the best-fit estimate.

How the three methods compare

Method	What it actually measures	Age estimate	Main uncertainty	Depends on ΛCDM?
CMB (Planck 2018)	Pattern of fluctuations 380,000 years after the Big Bang	13.797 ± 0.023 billion years	The cosmological model itself	Yes
Globular clusters	Main-sequence turnoff of the oldest star clusters	~12–13.5 billion years (plus formation time)	Stellar physics, distances	No
HD 140283	Distance and spectrum of one old star	14.5 ± 0.8 billion years (2013)	Distance, stellar models	No
Expansion rate (SH0ES H₀)	Present-day expansion + assumed composition	~12.7–12.8 billion years	Distance ladder systematics	Partly

The striking thing is not that the numbers are identical — they are not — but that methods based on nuclear physics inside stars, on the sound waves of the early plasma, and on the motions of nearby galaxies all land within about 10% of each other.

What "age of the universe" actually means

The age of the universe is the proper time elapsed since the Big Bang measured by a comoving observer — one moving with the average expansion of the universe, not moving peculiarly relative to it.

This is physically meaningful: it's the amount of time that has passed for matter that formed stars and galaxies from the Big Bang to now. It's not a statement about the size of the universe (which depends on whether the universe is finite or infinite, which is unknown).

The observable universe — the region from which light has had time to reach us — has a radius of about 46 billion light-years (not 13.8, because the universe has expanded during the travel time of that light). Our article on how big the universe is and how much of it we can see works through that geometry.

Cross-checks from the early universe

Several other observations do not measure the age directly but must be consistent with it:

• The earliest galaxies. In 2024 the James Webb Space Telescope confirmed JADES-GS-z14-0 at a redshift of 14.32, meaning we see it as it was less than 300 million years after the Big Bang. Galaxies that bright that early surprised many astronomers, but they do not contradict a 13.8-billion-year age; they constrain how fast the first stars formed. Our roundup of James Webb discoveries covers these finds.
• Big Bang nucleosynthesis. The amounts of helium, deuterium and lithium formed in the first minutes depend on the density of ordinary matter. The deuterium measurement agrees closely with the CMB, an independent check on the same model.
• Radioactive dating in the Solar System. Meteorites date the Solar System to about 4.57 billion years — comfortably younger than the universe, as it must be.

Before the Big Bang?

The Big Bang singularity is where general relativity predicts the breakdown of its own equations — a point of infinite density and temperature at t = 0. Whether "before the Big Bang" is a meaningful concept depends on quantum gravity, which we don't yet have. The age of the universe is really the age since the hot, dense, expanding phase began; it is silent on what, if anything, came before. Our explainer on how the Big Bang actually happened covers the first moments in more detail.

Some models:

• Eternal inflation: The hot Big Bang is the end of an inflationary epoch in our region; other inflationary regions exist "before" and "alongside" our bubble.
• Cyclic cosmology (Penrose's CCC): An infinite sequence of cosmic cycles, each starting with a hot dense state (like the Big Bang) and ending in a cold, expanded state that is rescaled into the next.
• No-boundary proposal (Hartle-Hawking): Near the Planck era, time behaves like a spatial direction; there is no "before" in any meaningful sense.

None are confirmed; none are ruled out.

Common misconceptions

"The Big Bang happened at a point, and we are measuring the time since that explosion." The Big Bang was not an explosion into pre-existing space; it was the rapid expansion of space everywhere. There is no centre we are measuring from.

"13.8 billion years is exact." It is the best fit within a specific model, with a statistical uncertainty of about 20–25 million years. The systematic uncertainty from the Hubble tension could be several hundred million years.

"If we see a galaxy 13 billion light-years away, it is 13 billion light-years away now." The light travelled for 13 billion years; the galaxy is now much farther away because space kept expanding.

"Stars older than the universe would disprove the Big Bang." It would signal an error in stellar models or cosmological parameters, but the Big Bang is supported by many other observations, including the CMB and light-element abundances.

FAQ

Q: How can the observable universe be 46 billion light-years in radius if the universe is only 13.8 billion years old? A: Light from the CMB has been travelling for about 13.8 billion years to reach us. But the region that emitted that light has continued to recede as the universe expanded. Today, that region is about 46 billion light-years away. We're seeing it as it was long ago, when it was much closer.

Q: Are there objects older than the universe? A: Not once uncertainties are included. HD 140283 was estimated at 14.5 ± 0.8 billion years in 2013, which overlaps with 13.8 billion years. If stellar ages clearly exceeded the cosmic age, it would indicate an error in either stellar physics or the cosmological model.

Q: Could the universe be much older? A: There is no strong evidence that it is dramatically older. The CMB, star ages and the expansion rate all constrain it fairly tightly. If the Hubble tension is resolved by new physics favouring a faster expansion, the universe would more likely be somewhat younger than 13.8 billion years, not older.

Q: What was the universe like when it was 1 second old? A: A hot, dense soup of protons, neutrons, electrons, photons and neutrinos at roughly 10 billion degrees. Protons and neutrons had already formed from quarks well before this point. Over the next few minutes, nucleosynthesis fused some of them into helium and traces of deuterium and lithium.

Q: What will cosmologists measure next to settle the Hubble tension? A: Key measurements include more gravitational-wave standard sirens (neutron star mergers measure distance without a ladder), JWST cross-checks of Cepheid distances, ground-based CMB experiments, and baryon acoustic oscillation maps from galaxy surveys such as DESI and Euclid.

Q: Why is 13.8 billion years quoted rather than 13.797? A: Because the extra digits imply a precision that depends on assuming the standard model is correct. Rounding to 13.8 billion years is honest about the real uncertainty, which is dominated by unresolved questions like the Hubble tension.

The bottom line

The universe is about 13.8 billion years old — a figure supported by independent lines of evidence that use different physics yet land within about 10% of each other. The CMB provides the most precise measurement but depends on the standard cosmological model; stellar ages confirm it from an entirely separate approach; and the expansion rate gives an estimate that is consistent but currently in tension with the CMB. That tension may turn out to be a measurement error or a sign of new physics. Either way, the evidence for a universe roughly 14 billion years old is among the most robust results in modern science.

Sources

• Planck Collaboration. Planck 2018 results. VI. Cosmological parameters. Astronomy & Astrophysics 641, A6 (2020).
• Riess AG et al. A Comprehensive Measurement of the Local Value of the Hubble Constant with 1 km/s/Mpc Uncertainty from HST and the SH0ES Team (2022).
• Freedman WL et al. Status Report on the Chicago-Carnegie Hubble Program (arXiv 2024; revised March 2025).
• ESA. Webb finds most distant known galaxy (2024).
• LIGO Scientific Collaboration. A gravitational-wave standard siren measurement of the Hubble constant. Nature 551, 85–88 (2017).

Related Reading

• How Big Is the Universe — and How Much of It Can We Actually See?
• The James Webb Space Telescope Explained — How It Works and What It's Found
• How Will the Universe End? Heat Death, Big Rip, and Big Crunch Explained
• What Is the Big Bang Theory? The Origin of the Universe Explained
• How Do We Actually Find Exoplanets?