What Is Dark Matter: Evidence, Candidates, and Open Questions

The Problem Dark Matter Solves

Dark matter was not invented by theorists in search of a puzzle — it was proposed reluctantly to explain observations that made no sense otherwise. The first major clue came from astronomer Fritz Zwicky in the 1930s, who measured the velocities of galaxies in the Coma cluster and found they were moving far too fast to be gravitationally bound by the visible matter alone. The galaxies should have flown apart long ago. Zwicky proposed an invisible mass he called dunkle Materie — dark matter — to account for the gravitational glue holding the cluster together.

The modern case for dark matter was built largely through the work of Vera Rubin and W. Kent Ford in the 1970s. Measuring the rotation curves of spiral galaxies, they found that stars in the outer regions of galaxies move at roughly the same speed as stars much closer to the center — exactly the opposite of what Newtonian gravity predicts if most of the mass is concentrated in the visible disk. A flat rotation curve implies that mass continues to exist far beyond where any light is detected, in a diffuse invisible halo surrounding the galaxy.

The Multiple Lines of Evidence

Dark matter is supported by several independent and mutually reinforcing lines of evidence, which makes the case qualitatively different from a single anomalous observation:

• Galaxy rotation curves: as described above, galaxies rotate too fast for their visible mass. This observation holds across thousands of galaxies.
• Gravitational lensing: light from distant galaxies is bent more than visible mass alone can account for, with the excess bending tracing a mass distribution that extends well beyond the visible galaxy.
• The Bullet Cluster: two galaxy clusters have collided and passed through each other. The hot gas (which interacts electromagnetically) was slowed by drag and forms a visible shock wave. The dark matter (which does not interact electromagnetically) passed straight through, creating a spatial offset between the center of mass (traced by gravitational lensing) and the center of visible matter. This is among the strongest direct evidence that dark matter is a real substance distinct from ordinary matter.
• Cosmic microwave background: the pattern of temperature fluctuations in the CMB depends sensitively on the composition of the early universe. Best-fit models require approximately 27% dark matter, 5% ordinary matter, and 68% dark energy.
• Large-scale structure: the web of galaxy filaments and voids visible in large galaxy surveys matches simulations only when dark matter is included. Without it, simulations produce structure that looks nothing like the observed universe.

What Dark Matter Is NOT

Several ordinary explanations have been tested and ruled out. Dark matter cannot be made of normal (baryonic) matter that is simply dark — dim stars, black holes, or cold gas — in sufficient quantities to explain the missing mass. Big Bang nucleosynthesis calculations constrain the total amount of baryonic matter in the universe to match the observed abundance of light elements (hydrogen, helium, lithium), leaving no room for an enormous additional reservoir of dark baryons. Searches for massive compact objects (MACHOs — massive astrophysical compact halo objects) using gravitational microlensing have also come up short.

Modifications to gravity — such as Modified Newtonian Dynamics (MOND) — can reproduce some galactic rotation curve data without dark matter but fail to explain cluster dynamics, the Bullet Cluster, and the CMB power spectrum simultaneously. No modified gravity theory has yet matched all the evidence that a particle dark matter model can explain.

Candidate Particles

The leading candidate for dark matter is a class of particles known as WIMPs (Weakly Interacting Massive Particles). WIMPs would be particles with masses roughly 1 to 1000 times the proton mass that interact with ordinary matter only through gravity and the weak nuclear force. Their appeal comes partly from the WIMP miracle: if WIMPs were produced thermally in the early universe (like other particle species), their relic density today would naturally come out near the observed dark matter density, without any fine-tuning.

Other serious candidates include axions — extremely light particles originally proposed to solve a separate problem in particle physics — and sterile neutrinos, which are hypothetical heavier cousins of the known neutrinos. More exotic proposals include primordial black holes (black holes formed in the early universe before any stars existed), though these are constrained by observations to account for at most a small fraction of dark matter.

Direct Detection Experiments

If WIMPs exist, they should occasionally collide with ordinary atomic nuclei, depositing a tiny amount of energy that could in principle be detected. A global network of ultra-sensitive detectors buried deep underground (to shield from cosmic rays) is searching for these collisions. Experiments including LUX-ZEPLIN (LZ) in South Dakota, XENONnT in Italy, and PandaX-4T in China use liquid xenon as a target medium cooled to cryogenic temperatures. None has detected a credible dark matter signal as of 2025.

Particle colliders — most notably the Large Hadron Collider (LHC) at CERN — could in principle produce dark matter particles directly if they are in the right mass range. No WIMP signal has been found at the LHC despite extensive searches. This has shifted the field somewhat toward lighter and more weakly coupled candidates, though WIMPs are not ruled out across their full mass range.

Indirect Detection and Astrophysical Searches

If dark matter particles can annihilate each other when they meet (as WIMPs are expected to), the annihilation should produce high-energy photons, neutrinos, or antimatter particles detectable by space-based telescopes. The Fermi Gamma-ray Space Telescope has searched for anomalous gamma-ray signals from the galactic center and from dwarf spheroidal galaxies (which are dark-matter-dominated). Some tentative anomalies have been reported but none has survived as a confirmed signal. The search continues with next-generation instruments.

The Open Questions

Despite overwhelming evidence that dark matter exists, its fundamental nature remains completely unknown. The failure to detect WIMPs after decades of increasingly sensitive searches has led some physicists to seriously consider lighter candidates (sub-GeV masses), stronger or weaker couplings than originally assumed, or entirely different classes of particles. The possibility that dark matter is not a single species but a rich dark sector with its own particles and forces is increasingly explored.

A definitive detection — whether through direct collision, production at a collider, or unambiguous astrophysical signal — would be one of the most significant scientific discoveries in history. The search represents one of the most active and well-funded frontiers in fundamental physics, spanning underground laboratories, space telescopes, and the world's most powerful particle accelerators simultaneously.