What is the evidence for dark matter's existence?

Five converging pieces of evidence: 1) flat rotation curves of galaxies (Vera Rubin, 1970); 2) velocity dispersions in galaxy clusters (Zwicky, 1933); 3) gravitational lensing, where light is deflected by invisible mass; 4) the cosmic microwave background (CMB) mapped by WMAP and Planck, which requires a precise dark matter density; 5) the formation of large-scale cosmic structures (filaments, voids), only explained with a non-baryonic gravitational component. The Bullet Cluster (2006) remains one of the most spectacular observations: visible matter and dark matter are spatially separated after a collision.

What fraction of the universe is dark matter?

According to the most precise measurements from ESA's Planck satellite (2018 final results), the universe is composed of: 68.3% dark energy, 26.8% dark matter, and only 4.9% ordinary (baryonic) matter — stars, planets, gas, you and me. In other words, the 'normal' matter described by all of chemistry, biology and particle physics accounts for only about 5% of the universe. Of the remaining 95%, dark matter makes up more than a quarter, and its nature remains completely unknown.

What is dark matter?

Q: Why is it called 'dark' matter?

The term 'dark' here means 'invisible to telescopes', not 'dark' in the optical sense. Dark matter neither emits, reflects nor absorbs light (nor any other electromagnetic waves: radio, infrared, X-rays, gamma rays). It is therefore strictly transparent. The word 'dark' designates our inability to detect it directly with conventional optical instruments. Its presence is only revealed indirectly, through its gravitational effects on visible matter and on light.

Q: How was dark matter discovered?

Swiss astronomer Fritz Zwicky proposed the concept as early as 1933 while studying the Coma cluster. But the decisive evidence came in 1970 with American astronomer Vera Rubin and her collaborator Kent Ford. By measuring the rotation speed of stars in spiral galaxies, they discovered that stars far from the center rotate as fast as those closer in, contradicting Newton's laws if only visible matter exists. To explain these 'flat rotation curves', one must postulate an invisible halo with about 5 to 10 times more mass than the visible stars.

Q: Is dark matter made of particles?

The dominant hypothesis is that dark matter is made of unknown massive particles, interacting only through gravity and possibly the weak nuclear force. The Standard Model of particle physics contains no valid candidate: neutrinos, electrons, quarks and photons are all ruled out. Physics 'beyond the Standard Model' must therefore be invoked. The main theoretical candidates are WIMPs, axions, sterile neutrinos, and supersymmetric particles.

Q: What are WIMPs and axions?

WIMPs (Weakly Interacting Massive Particles) are hypothetical massive particles (10 to 1000 times the proton mass) interacting only through gravity and the weak nuclear force. They are predicted by supersymmetric theories. Axions are ultra-light particles (10⁻⁶ to 10⁻³ eV) initially proposed in 1977 by Roberto Peccei and Helen Quinn to solve a symmetry problem in quantum chromodynamics. Both candidates remain the most actively studied in 2026, but neither has been directly detected to date.

Q: Why have XENON and LUX experiments not detected it?

Direct detection experiments (XENONnT in Italy, LUX-ZEPLIN in the United States, PandaX in China) use enormous tanks of liquid xenon buried under kilometers of rock to block cosmic rays. They look for the recoil of an atomic nucleus struck by a WIMP. Since 2010, these experiments have continually increased their sensitivity without detecting anything. This means either WIMPs do not exist, or their cross-section is much smaller than predicted, or dark matter is made of axions or another category. The accumulated null results exclude a large portion of the supersymmetric parameter space.

Q: What is the link with ordinary (baryonic) matter?

Dark matter and baryonic matter are two distinct components. Baryonic matter (protons, neutrons, electrons) interacts with light and makes up all 'visible' matter: stars, planets, interstellar gas. Dark matter does not interact with light, but they share gravitational attraction. Cosmological simulations show that baryonic matter concentrates in the potential wells created by dark matter: without dark matter, galaxies probably would not have had time to form since the Big Bang.

Q: Does dark matter interact with itself?

Most models assume dark matter is 'cold' (CDM, Cold Dark Matter) and interacts with itself only through gravity. But certain small-scale anomalies (the missing satellites problem, the density profile at the center of dwarf galaxies) have led to proposals of self-interacting dark matter (SIDM), which could have its own invisible forces. The Bullet Cluster observation imposes strict limits, however: if dark matter strongly self-interacted, it would have decelerated during the collision, which is not observed.

Q: Could it be a misunderstanding of gravity?

This is the hypothesis behind MOND (Modified Newtonian Dynamics), proposed in 1983 by Israeli physicist Mordehai Milgrom. MOND modifies Newton's law at very low accelerations and remarkably well explains galactic rotation curves without dark matter. However, MOND fails to explain the Bullet Cluster, the cosmic microwave background and the formation of large-scale structures. Its relativistic version (TeVeS) is now widely considered insufficient. The majority of cosmologists in 2026 still favor the particle hypothesis, while acknowledging that MOND captures something real about individual galaxies.

A rigorous definition, observational evidence, theoretical candidates (WIMPs, axions), detection experiments (XENON, LUX-ZEPLIN) and alternatives (MOND): everything we know — and don't know — about the matter that constitutes 27% of the universe.

TL;DR — Quick answer

Dark matter is an invisible form of matter that makes up about 27% of the universe. It neither emits nor absorbs light, but its presence is revealed through its gravitational effects on galaxies. Its exact nature remains one of the greatest mysteries of modern physics.

Vigi-Sky Team · May 8, 2026 · 12 min read · Reference cosmology

Why is it called "dark" matter?

The term "dark" here does not mean "obscure" in the optical sense: it means invisible to instruments. Dark matter neither emits, reflects nor absorbs light, nor any other electromagnetic wave — radio, infrared, X-rays, gamma rays. It is, literally, transparent.

This total invisibility is its defining property. A star shines, an interstellar gas cloud emits in infrared, a black hole betrays its presence by the X-ray jets of its accretion disk. Dark matter does none of these. Its only signature is gravitational.

Dark Matter: Invisible, non-baryonic component whose presence is revealed solely through its gravitational effects.
Dark Energy: Not to be confused with dark matter. The component responsible for the accelerating expansion of the universe (about 68%). Completely different nature.
Baryonic Matter: "Ordinary" matter made of protons and neutrons: stars, planets, gas, you and me (about 5% of the universe).

How was dark matter discovered?

The first hint came in 1933 from Swiss astronomer Fritz Zwicky. While studying the Coma cluster of galaxies, he calculated that the galaxies there move too fast: they should escape if only visible matter existed. He proposed the existence of dunkle Materie ("dark matter"). His hypothesis was largely ignored for nearly 40 years.

The decisive proof came in 1970 with the work of American astronomer Vera Rubin and her collaborator Kent Ford. They measured the rotation speed of stars in the Andromeda galaxy (M31), then in dozens of other spiral galaxies.

In a spiral galaxy, the ratio of dark-to-light matter is about a factor of ten. That's probably a good number for the ratio of our ignorance to knowledge. — Vera Rubin, Carnegie Institution, 2000

According to Newton's laws, stars far from the galactic center should rotate more slowly (like distant planets around the Sun). But Rubin observed the opposite: the rotation curves are flat. To explain this, one must postulate an enormous invisible halo with 5 to 10 times the mass of the visible stars.

What is the evidence for its existence?

Five independent pieces of evidence converge to support dark matter:

Flat rotation curves of spiral galaxies (Vera Rubin, 1970).
Velocity dispersions in galaxy clusters (Zwicky 1933, since confirmed).
Gravitational lensing: light from distant galaxies is bent by invisible mass, mappable with Hubble and JWST.
Cosmic microwave background (CMB): WMAP (2003-2010) and Planck (2013-2018) measured the precise imprint of dark matter in the Big Bang afterglow.
Large-scale structure formation: the cosmic filaments observed (Sloan Digital Sky Survey, DESI) only formed in 13.8 billion years thanks to a non-baryonic gravitational "glue".

The most spectacular observation remains the Bullet Cluster (NASA, 2006): two colliding galaxy clusters where visible matter (X-ray gas) and dark matter (mapped by gravitational lensing) are spatially separated. This result is extremely difficult to explain without dark matter, and has been called the "smoking gun" by cosmologists.

What fraction of the universe does dark matter represent?

68.3 %

Dark energy

26.8 %

Dark matter

4.9 %

Ordinary matter

≈ 0.1 %

Visible stars

According to the most precise measurements from ESA's Planck satellite (final 2018 results), the universe consists of 68.3% dark energy, 26.8% dark matter, and only 4.9% ordinary (baryonic) matter. In other words, the "normal" matter described by all of chemistry, biology and particle physics accounts for only 5% of the universe. Of the remaining 95%, dark matter makes up more than a quarter, and its nature remains completely unknown.

Is dark matter made of particles?

The dominant hypothesis is that dark matter is made of unknown massive particles, interacting only through gravity (and possibly the weak nuclear force). The Standard Model of particle physics contains no valid candidate: neutrinos, electrons, quarks, photons are all ruled out for reasons of mass, stability, or abundance.

Physics "beyond the Standard Model" must therefore be invoked. The main theoretical candidates currently being explored are WIMPs, axions, sterile neutrinos and supersymmetric particles.

What are WIMPs and axions?

WIMPs (Weakly Interacting Massive Particles): Hypothetical massive particles (10 to 1000 times the proton mass) interacting only through gravity and the weak nuclear force. Predicted by supersymmetric (SUSY) theories. The dominant "historical" candidate of the 1990s-2010s.
Axions: Ultra-light particles (10⁻⁶ to 10⁻³ eV) initially proposed in 1977 by Roberto Peccei and Helen Quinn to solve a symmetry problem in quantum chromodynamics (QCD). Do not require supersymmetry. Searched for by the ADMX and IAXO experiments.
Sterile neutrinos: Hypothetical fourth family of neutrinos interacting only through gravity. Searched for notably by the eROSITA X-ray telescope.
MACHOs (Massive Compact Halo Objects): Hypothesis now largely abandoned: brown dwarfs, ancient white dwarfs, primordial black holes. Gravitational microlensing surveys (EROS, MACHO, OGLE) have ruled this out for the bulk of the missing mass.

Why have the XENON and LUX experiments not detected it?

Direct detection experiments — XENONnT (Italy, Gran Sasso Laboratory), LUX-ZEPLIN (USA, Sanford Lab), PandaX-4T (China, Jinping Lab) — use enormous tanks of ultra-pure liquid xenon buried under kilometers of rock to block cosmic rays. They look for the tiny recoil of an atomic nucleus struck by a WIMP from the galactic halo.

Since 2010, these experiments have continually increased their sensitivity without detecting anything. This does not mean the absolute failure of the hypothesis: it means either WIMPs do not exist, or their cross-section is much smaller than predicted, or dark matter is made of axions or another totally different category.

The accumulated null results in 2026 exclude a large portion of the supersymmetric parameter space. Experiments are now approaching the "neutrino floor", a threshold beyond which solar neutrinos become an irreducible background.

What is the link with ordinary (baryonic) matter?

Dark matter and baryonic matter are two distinct components. Baryonic matter (protons, neutrons, electrons) interacts with light and makes up all "visible" matter: stars, planets, interstellar gas. Dark matter does not interact with light, but they share gravitational attraction.

Cosmological simulations (Millennium Simulation, IllustrisTNG, EAGLE) show that baryonic matter concentrates in the potential wells dug by dark matter. Without dark matter, galaxies probably would not have had time to form since the Big Bang: baryonic fluctuations were too weak, smoothed by their coupling with photons until recombination (380,000 years after the Big Bang).

Key takeaway

Dark matter is the gravitational scaffolding upon which all galaxies — including our own — formed. Without it, the Milky Way would probably not exist.

Does dark matter interact with itself?

Most models assume dark matter is "cold" (CDM, Cold Dark Matter) and interacts with itself only through gravity. But certain small-scale anomalies — the "missing satellites problem", the density profile at the core of dwarf galaxies — have led to proposals of self-interacting dark matter (SIDM, Self-Interacting Dark Matter), which could have its own invisible forces.

The Bullet Cluster observation imposes strict limits, however: if dark matter strongly self-interacted, it would have decelerated during the collision, like the X-ray gas. But it is observed to have passed through unimpeded. Self-interaction cross-sections must therefore be very small.

Could it be a misunderstanding of gravity?

This is the hypothesis behind MOND (Modified Newtonian Dynamics), proposed in 1983 by Israeli physicist Mordehai Milgrom. MOND modifies Newton's law at very low accelerations (below about 10⁻¹⁰ m/s²) and remarkably well explains galactic rotation curves without any dark matter.

However, MOND fails to explain several crucial observations:

The Bullet Cluster and the separation of visible matter from gravitational mass.
The precise acoustic peaks of the cosmic microwave background measured by Planck.
The formation of large-scale cosmic structures at scales of hundreds of megaparsecs.
The behavior of galaxy clusters as a whole.

Its relativistic version (TeVeS, Bekenstein 2004) is now widely considered insufficient. The majority of cosmologists in 2026 still favor the particle hypothesis, while acknowledging that MOND captures something real about individual galaxies. The debate remains open and fascinating.

Key quote

"The night sky contains far more mass than stars. Most of the universe is, literally, invisible to us." — adapted from Vera Rubin's analysis of rotation curves.