Home / Learn / Science / Cosmology A to Z / Lesson 8.8
All lessons

Module 8: Dark matter · Lesson 8.8 · about 13 min

Dark matter: what we know and what we don't

Dark matter is 85% of all matter: cold, collisionless and not atoms, seen on every scale from dwarf galaxies to the CMB. What it is made of is still unknown.

Key points, meanings and formulas on one page: choose "Save as PDF" in the print box.

Module 8 · about 13 minutes. Read first: 8.7 MOND and friends. This lesson gathers the module together: the evidence on every scale, what we can say for sure about dark matter, and the questions still open. By the end you will be able to sum up the case for dark matter in a minute, and say honestly what nobody knows yet.

Dark matter is one idea that explains measurements on every scale, from small galaxies to the whole observable universe, with one set of numbers. What it is made of is one of the biggest open questions in science.

1. The evidence, scale by scale

Galaxies, tens of kiloparsecs. Flat rotation curves need dark halos several times heavier than the stars (Lesson 8.2). Both cold dark matter and MOND fit them; MOND does it with one number and no halo.
Clusters, a few megaparsecs. Galaxy speeds, hot gas and lensing all weigh the same mass: about 85% dark (Lesson 8.1). MOND still needs about twice the visible mass.
The Bullet Cluster. Lensing mass went with the galaxies, away from the gas that holds most of the atoms (Lesson 8.3). Easy with dark matter; very hard with modified gravity alone.
The cosmic web, tens to hundreds of megaparsecs. Weak lensing surveys map dark matter directly, and galaxy surveys find the web of filaments and voids that cold dark matter simulations predict (Module 10).
The first light, the whole sky. The CMB peaks weigh dark matter and atoms separately, 5.4 to 1, and atoms alone could not have grown galaxies in time (Lesson 8.4). Nucleosynthesis agrees: dark matter is not atoms.
And the puzzles. On the smallest scales, dwarf galaxies do not always match simple dark matter simulations. Most of these puzzles are easing as simulations include the effects of stars and gas, but they are where new physics could still hide.

2. What we know

  • How much: Ωc ≈ 0.26, about 5.4 times the atoms and about 84% of all matter, known to about 1%.
  • Where: in halos round every galaxy and cluster, and along the filaments of the cosmic web (Lesson 10.3). Near the Sun about 0.3 to 0.5 GeV per cubic centimetre.
  • What it is not: not atoms, not ordinary neutrinos, not mostly dim stars or stellar black holes.
  • How it behaves: it feels gravity; it is cold or at most warm (it makes small galaxies); it hardly collides with itself (less than about 1 cm² per gram) or with atoms; it has lasted 13.8 billion years.

How thin is it? On average across the universe, dark matter is only about 1.3 times the mass of a proton per cubic metre. In our halo it is hundreds of thousands of times denser than that, yet still only about a kilogram is passing through the whole Earth at any moment.

3. What we don’t know

  • What it is. WIMPs, axions, sterile neutrinos, fuzzy dark matter, primordial black holes, or something nobody has thought of. Its mass is uncertain by about 90 powers of ten (Lesson 8.5).
  • Whether it feels anything but gravity. Every search so far, direct, indirect and at colliders, has come up empty (Lesson 8.6). It may interact only through gravity, which would make it very hard to catch in a laboratory.
  • Whether it is one thing. Ordinary matter is protons, neutrons, electrons, photons and neutrinos. A “dark sector” could be just as varied.
  • The small-scale puzzles. Simulations of cold dark matter alone predict dense “cusps” in the middles of small galaxies, while many dwarfs have smoother “cores”; they predicted more small satellite galaxies than were known; and some predicted satellites are too dense compared with the real ones (“too big to fail”). Supernova winds can stir up the middles of small galaxies, and dozens of faint new dwarfs have been found since 2005, so these are now seen as tests of how galaxies form, or hints of warm, fuzzy or self-interacting dark matter, rather than failures.
  • The tensions. Lumpiness measured by weak lensing comes out a little lower than the CMB predicts (Lesson 12.3), and JWST finds bright galaxies surprisingly early (Lesson 12.4). Both test the cold dark matter picture.

4. What comes next

Bigger xenon detectors, such as the planned XLZD with about 60 tonnes, will push towards the neutrino fog. Axion experiments are speeding up their scans of new mass ranges. The Euclid space telescope (launched 2023) and the Vera C. Rubin Observatory (first images 2025) will map dark matter by weak lensing across billions of galaxies, and NASA’s Roman Space Telescope will join them. Each could find a crack in the picture, or the particle itself.

Ninety years in one line. Zwicky’s clusters (1933), Rubin’s and Bosma’s rotation curves (1970s), cold dark matter theory (1982-85), the CMB’s weighing (2003 onwards), the Bullet Cluster (2006), and the great underground searches (2010s and 2020s). The evidence has grown steadily stronger; the identity has stayed hidden.

“Dark matter and dark energy are the same thing, both just mean ‘we don’t know’.” They are opposites in what they do. Dark matter clumps and pulls, holding galaxies and clusters together, and thins out as the universe grows. Dark energy is smooth, does not thin out, and makes the expansion speed up (Module 9). They share only the word “dark”, meaning we cannot see them.

  • Dark matter explains rotation curves, clusters, lensing, the Bullet Cluster, the cosmic web and the CMB with one set of numbers.
  • It is about 26% of the universe and 84% of all matter: cold, nearly collisionless, stable and not made of atoms.
  • Its identity is unknown; no experiment has detected it except through gravity.
  • Small-scale puzzles and modified gravity remain active questions, but cold dark matter is the mainstream picture.

Check yourself

1Roughly what fraction of all matter is dark matter?
2Which of these do we NOT yet know about dark matter?
3How is dark matter different from dark energy?

Sources

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.