What the universe is made of
A working reference on the evidence, candidates, and open questions around dark matter — the unseen mass inferred from its gravity but not yet directly detected. Every figure below traces to a numbered source in the provenance panel.
How we know something is there
Four independent lines of evidence — galactic rotation curves, gravitational lensing, the cosmic microwave background, and the growth of large-scale structure — converge on the same conclusion: roughly five-sixths of the universe's matter does not emit, absorb, or scatter light. None of these methods detects a particle directly; each infers mass from its gravitational effect. That an unseen mass exists is essentially uncontested; what it is made of is the open question this console tracks.
Observational evidence
Five independent measurements, each pointing to mass that isn't visible.
Galaxy rotation curves
consensusStars and gas orbiting the outer edges of spiral galaxies move at roughly constant speed with distance from the center, rather than slowing down as Kepler's laws predict for mass concentrated in the visible disk. Vera Rubin and Kent Ford's spectroscopic surveys of Andromeda first showed this pattern; the modern SPARC database extends it across more than 150 galaxies. A minority of researchers instead attribute flat rotation curves to modified gravity (MOND) rather than unseen mass; the lensing and CMB evidence below are harder for that view to accommodate on its own.
Gravitational lensing — the Bullet Cluster
consensusIn the merging cluster system 1E 0657-558, weak-lensing maps show the bulk of the gravitational mass spatially offset from the hot X-ray-emitting gas — the dominant visible mass component — because the gas clouds collided and slowed while the (mostly collisionless) mass passed through unimpeded. Clowe et al. reported this offset at 8σ significance, a result that is difficult to reproduce with a modified gravity law and no unseen mass.
Cosmic microwave background acoustic peaks
consensusThe relative heights of the acoustic peaks in the CMB's temperature power spectrum depend on the ratio of matter that clumps gravitationally without interacting with light (cold dark matter) to matter that does interact with light (baryons). The Planck satellite's full-mission data fix that ratio to sub-percent precision, within the standard ΛCDM model.
Large-scale structure growth
consensusOrdinary matter alone couples to radiation pressure too strongly to have clumped into galaxies and clusters within the age of the universe. Simulations require a gravitationally dominant, non-relativistic (“cold”) dark matter component to seed structure early enough to match observed galaxy clustering — a picture consistent with the Planck-derived density parameters.
Big Bang nucleosynthesis
consensusThe observed abundances of light elements (helium, deuterium, lithium) constrain the density of ordinary (baryonic) matter independently of the CMB. That baryon density falls far short of the total matter density needed to explain rotation curves and structure formation, meaning most of the missing mass cannot be ordinary matter in any form — it must be something non-baryonic.
Candidate particles and objects
What could the missing mass actually be? None of these is confirmed; each has a distinct experimental signature.
WIMPs — weakly interacting massive particles
under direct testHypothetical particles in the roughly 1 GeV–10 TeV mass range that interact via gravity and the weak force. Attractive because a thermal relic with weak-scale interactions naturally freezes out at close to the observed dark matter abundance (the “WIMP miracle”). Directly searched for in underground xenon detectors — see the Direct Detection panel. No confirmed signal after two decades of increasingly sensitive searches, which has narrowed but not closed the viable parameter space.
Axions and axion-like particles
speculativeExtremely light (µeV–meV scale) particles originally proposed to solve an unrelated problem in the strong nuclear force (the strong CP problem), which would also work as cold dark matter if produced in the early universe. Searched for via their predicted weak coupling to photons in strong magnetic fields, an approach pursued by experiments such as ADMX. Large parts of the theoretically preferred mass range remain unexplored.
Sterile neutrinos
speculativeA hypothetical keV-scale neutrino that would not interact via the weak force at all, evading ordinary neutrino detection methods. Would behave as “warm” dark matter — free-streaming enough to slightly suppress structure on small scales, a signature that current small-scale structure observations constrain but have not ruled out entirely.
Primordial black holes
speculative, narrowly constrainedBlack holes that could have formed from density fluctuations in the first second after the Big Bang, rather than from stellar collapse. Could in principle be made entirely of ordinary matter and still count as “dark.” Microlensing surveys, CMB distortions, and gravitational-wave merger-rate observations have excluded most possible mass ranges as the dominant dark matter component, though narrow mass windows remain open.
MACHOs — massive compact halo objects
disfavored as the dominant componentOrdinary-matter objects — brown dwarfs, rogue planets, black holes — in the mass range a stellar telescope wouldn't easily see. Large microlensing surveys of the Magellanic Clouds (the MACHO and EROS-2 collaborations) searched for these objects passing in front of background stars and found far too few events to account for a galaxy's dark halo, effectively ruling MACHOs out as the primary explanation across most plausible mass ranges.
Direct detection
Underground detectors look for a dark matter particle scattering off an atomic nucleus. No confirmed signal has ever been seen — every published result to date is an exclusion limit, ruling out interaction strengths above a given line.
2.1 × 10⁻⁴⁸ cm²
at 90% confidence, for a WIMP mass of 36 GeV/c², set by the LUX-ZEPLIN (LZ) experiment's 2024 analysis of 4.2 tonne-years of exposure at the Sanford Underground Research Facility. XENONnT and PandaX-4T report comparable sensitivity in the same mass range using the same liquid-xenon technique.
Cosmological parameters
The values below are the Planck Collaboration's 2018 final full-mission cosmic microwave background analysis, assuming the standard flat ΛCDM model.
| Parameter | Value | Meaning |
|---|---|---|
| Ωch² | 0.120 ± 0.001 | Cold dark matter density |
| Ωbh² | 0.0224 ± 0.0001 | Baryon (ordinary matter) density |
| H₀ | 67.4 ± 0.5 km/s/Mpc | Present expansion rate (from the CMB) |
| Ωm | 0.315 ± 0.007 | Total matter density fraction |
| ΩΛ | 0.685 | Dark energy density fraction (flat universe) |
| σ₈ | 0.811 ± 0.006 | Amplitude of matter clustering |
Provenance & governance
Every figure on this console traces to a numbered source below. This panel also records how the dataset is maintained.
Editorial policy
Figures are drawn from peer-reviewed collaborations (Planck, LZ) and landmark peer-reviewed papers, cited by author, venue, and year. Confidence tiers (see legend on the Evidence and Candidates panels) distinguish results with broad scientific consensus from single-experiment or model-dependent claims, and from genuinely speculative or excluded ideas. Illustrative chart curves are explicitly labeled as schematic rather than extracted data. Contested empirical questions — modified gravity as an alternative to particle dark matter, the Hubble tension — are flagged rather than silently resolved. This console does not represent original research; it summarizes published results and will contain the simplifications inherent to any summary.
Sources
Changelog
| Version | Date | Change |
|---|---|---|
| v1.0 | 2026-09-01 | Initial publication: overview, five evidence lines, five candidate profiles, LZ 2024 detection limit, Planck 2018 parameter table. |
Known limitations
This console simplifies for a general audience: exclusion-limit curves are schematic rather than extracted from published data files; candidate mass ranges are given as representative orders of magnitude rather than exact bounds, which shift with each new experimental run; and the modified-gravity minority position is noted but not developed in full. Treat this as an orientation layer, not a substitute for the primary sources above.
