The universe has no birth certificate. The estimate of roughly 13.8 billion years comes from fitting a physical history to observations, then checking that history against older stars. It is one of astronomy's most strikingly precise results, although the precision belongs to a model as well as to the instruments.¹
Run expansion backwards, carefully
Light from distant galaxies is stretched as space expands. Astronomers measure that stretching as redshift and compare it with distance. Nearby distance estimates are built in steps: methods such as Cepheid variable stars calibrate the brightness of more distant markers, including certain supernovae. The result is a measurement of today's expansion rate, the Hubble constant.²
A first guess might be to divide distance by speed and run the cosmic film backwards. The inverse of the Hubble constant gives an age scale, but not the full age. Expansion has not proceeded at a fixed rate. Matter's gravity slowed it earlier; the component called dark energy is associated with its later acceleration. To turn a present-day rate into elapsed time, astronomers must describe how the rate changed.¹,²
That distinction is more than a technical footnote. Imagine two cars now travelling at the same speed. If one accelerated from rest and the other spent most of its journey slowing down, their current speeds alone cannot tell you when they set out. Cosmic expansion also needs a history, not just a snapshot.
Read the oldest available light
About 380,000 years after the hot early universe began expanding, it cooled enough for light to travel freely. That radiation has been stretched into microwaves and now reaches us from every direction as the cosmic microwave background, or CMB. It is not light from the instant of the Big Bang; it is the earliest light we can directly observe.³
The CMB is almost uniform, but it contains tiny temperature variations. Their pattern records sound waves and density differences in the early universe. Satellites such as WMAP and Planck mapped those variations across the sky. A cosmological model predicts what patterns should appear for different amounts of ordinary matter, dark matter and dark energy, and for different expansion histories. Fitting the observed pattern constrains those ingredients.⁴
Once the ingredients and expansion history are specified, the equations can be integrated from the early hot state to today. The Planck collaboration's standard six-parameter model, combined with other observations in one published fit, gives an age of 13.787 billion years with a quoted statistical uncertainty of 0.020 billion years.⁵ That is about twenty million years, tiny beside the total age. The precise decimals are a property of that particular model and data combination, not a direct stopwatch reading.
Why can an image of faint microwaves carry so much information? It contains many independent-looking features across the sky, measured in temperature and polarisation, whose sizes and strengths respond in different ways to the same underlying parameters. A model that moves one parameter to fit a feature can spoil its fit to others. Instrument calibration and foreground light from our own galaxy also have to be modelled, so more data do not eliminate every systematic uncertainty.⁴,⁵
Let ancient stars check the clock
Stars cannot be older than the universe that formed them. That makes the oldest stars a largely different kind of test. In a globular cluster, many stars formed at roughly the same time. Massive stars burn their fuel faster and leave the ordinary sequence of hydrogen-burning stars sooner. The location of this “turn-off” in a cluster's pattern of brightness and colour helps astronomers estimate its age.¹,⁶
White dwarfs offer another clock: these stellar remnants cool over time. Their faintness and temperature can constrain how long their parent population has existed. Such methods need accurate distances, chemical compositions and models of stellar interiors. They are not simple readings from a star's surface.⁶
A recent globular-cluster analysis using Gaia distances estimated a universe age near 13.57 billion years, with statistical and additional systematic uncertainties.⁷ The result is consistent with the CMB estimate once those uncertainties are included. An old cluster provides a lower bound on cosmic age; adding the time between the Big Bang and the cluster's formation turns it into a fuller estimate. The uncertainty in that formation delay matters.
This cross-check has a useful logical shape. If well-established stellar physics produced stars unambiguously older than a well-established cosmic model, something would be wrong. Agreement does not prove every assumption, but it makes a large error harder to hide.
A useful scale check is the Solar System. Its roughly 4.6-billion-year age, found from radiometric dating of ancient material, is far younger than the cosmic estimate.³ That leaves ample time for earlier generations of stars to make heavier elements before the Sun formed. This is not a separate high-precision cosmic clock, but it is another test of the order of events: a proposed universe younger than the Solar System would fail immediately.
What is still unsettled?
Measurements of the nearby universe's current expansion rate remain in tension with the value inferred from the CMB under the standard model. This “Hubble tension” is about how those methods relate, not evidence that astronomers have two directly measured birth dates. It could reflect an unrecognised systematic error, incomplete cosmological physics or some combination. NASA reports that improved distance measurements have not yet removed the puzzle.²,⁸
The answer, then, is precise with an asterisk that deserves to be read. Under the model that best fits a broad set of observations, the universe is about 13.8 billion years old. The oldest stars provide an independent reality check. Future measurements may adjust the detailed figure or the model behind it; they are unlikely to turn the cosmic clock into a matter of guesswork.