Module 4: Discovering expansion · Lesson 4.7 · about 14 min
How old is the universe?
How we know the universe is about 13.8 billion years old: the Hubble time 1/H0, why the history of the expansion changes the answer, the 1990s age problem when stars seemed older than the universe, and three independent checks that now agree.
Module 4 · about 14 minutes. Read first: 4.6 Lemaître and the primeval atom. Rewinding the expansion points to a beginning. Now we put a date on it. By the end you will be able to estimate the age of the universe from H0 and explain why the history of the expansion matters.
1. The Hubble time
If galaxies had always moved apart at today’s speeds, the time since they were together would be distance ÷ speed. With v = H0 d, that is d ÷ (H0 d) = 1/H0, the same for every galaxy.
- tH
- the Hubble time: the age if the expansion had never sped up or slowed down
- 977.8
- the number that turns km/s per Mpc into billions of years (1 Mpc ÷ 1 km/s ≈ 977.8 billion years)
- Physical meaning
- A faster expansion today means less time to reach today's distances: a younger universe. H0 = 70 gives 14.0 billion years.
2. Slowing down and speeding up
Real expansion did not run at a steady pace. Gravity of matter slowed it at first, so galaxies moved faster in the past and reached today’s distances sooner: the true age is less than 1/H0. A universe of matter alone would be exactly 2/3 of its Hubble time old. Dark energy, which began to speed the expansion up about 6 billion years ago, has been doing so ever since, which makes the age longer.
3. The age problem of the 1990s
Ages of the oldest stars come from globular clusters, ancient balls of hundreds of thousands of stars, all born together. How bright and hot their stars are, compared with models of how stars burn, shows they are roughly 12 to 13.5 billion years old.
In the early 1990s most cosmologists expected a universe of matter only, and measurements of H0 were drifting towards 70 to 80. That gave an age of only 8 to 9 billion years: younger than the stars inside it, which is impossible. Either H0 was wrong, the star models were wrong, or the universe was not made of matter alone.
Solved by dark energy. In 1998 two teams found from supernovae that the expansion is speeding up (Module 9). A universe with dark energy is older for the same H0. With the Hubble Key Project’s H0 of 72 in 2001, the age came out at about 13 to 14 billion years, and in 2003 the WMAP satellite pinned it at 13.7: older than the oldest stars.
4. Three clocks that agree
- Expansion history: the Planck satellite's map of the cosmic microwave background, read with the standard model, gives 13.79 ± 0.02 billion years (2018).
- Oldest stars: globular clusters, about 12 to 13.5 billion years, and they formed some time after the start.
- Radioactive dating: uranium and thorium in very old stars decay like clocks and give similar ages, with bigger error bars.
Three methods that share no assumptions agree. That is why cosmologists are confident in “13.8 billion years”. The Hubble tension (Lesson 4.2) moves the answer by less than a billion years either way.
“The universe is 13.8 billion years old, so the edge of what we can see is 13.8 billion light-years away.” Space has been stretching while the light travelled, so the farthest regions we see are now about 46 billion light-years away. Lesson 5.9 explains how.
- The Hubble time 1/H0 = 977.8 ÷ H0 billion years (about 14 billion for H0 = 70) is a first estimate of the age.
- Slowing by gravity makes the age shorter (matter only: 2/3 of 1/H0); speeding up by dark energy makes it longer.
- In the 1990s matter-only models gave ages younger than the oldest stars; dark energy solved this.
- The CMB with the standard model gives 13.8 billion years; globular clusters and radioactive dating agree.
Work out the age of the universe at any redshift with the [[cosmology-calculator]].
Check yourself
Sources
- Planck Collaboration (2020), Planck 2018 results. VI. Cosmological parameters, A&A 641, A6 https://doi.org/10.1051/0004-6361/201833910
- D. Valcin et al. (2020), Inferring the age of the universe with globular clusters, JCAP 2020 (12), 002 https://doi.org/10.1088/1475-7516/2020/12/002
- W. L. Freedman et al. (2001), Final results from the Hubble Space Telescope Key Project to measure the Hubble constant, ApJ 553, 47 https://doi.org/10.1086/320638
Your ticks stay in this browser. Save them as a file on the course home.
Updated 2026-11-22. Numbers are rounded for teaching; measured values come from the sources listed in each lesson. Simulations run in your browser; nothing you type is sent anywhere.