Cosmology keeps shrinking the neutrino mass

The neutrino is the lightest particle of matter we know of, and for nearly a century physicists could not weigh it at all. Now the problem is almost the opposite: the most powerful cosmological surveys keep pushing the summed mass of the three neutrinos down, year after year, until the number has become so small that it is bumping into a hard floor set by an entirely different kind of experiment. The latest data place the sum below about 0.064 electronvolts — and a careful statistical treatment pushes it lower still, into territory that neutrino oscillation measurements say is forbidden. Something in the picture has to give.

A bound that keeps falling

Cosmology weighs neutrinos indirectly. Because they carry a tiny but non-zero mass, neutrinos streaming through the early universe smoothed out the clumping of matter on certain scales, leaving a measurable imprint on the distribution of galaxies and on the cosmic microwave background. The heavier the neutrinos, the more smoothing — so a map of cosmic structure is, in effect, a scale. (We unpacked exactly how this works in how cosmology weighs the neutrino.)

For years that scale read “less than about 0.12 eV.” Then the Dark Energy Spectroscopic Instrument, or DESI, began delivering the largest three-dimensional map of the universe ever made. Combining DESI’s Data Release 2 measurements of baryon acoustic oscillations with cosmic microwave background data from Planck and the Atacama Cosmology Telescope, the collaboration reported an upper limit on the summed neutrino mass, Σm_ν, of just 0.064 eV at 95 percent confidence, assuming the standard ΛCDM cosmological model. That is roughly seven times tighter than the best direct laboratory limit from KATRIN, and it is closing fast.

The collision with oscillations

Here is why that number is uncomfortable. Oscillation experiments — the ones that earned the 2015 Nobel Prize — do not measure neutrino masses directly, but they pin down the differences between the squared masses with great precision. Those differences impose a floor: the three masses cannot sum to less than about 0.059 eV if they follow the so-called normal ordering, or about 0.10 eV for the inverted ordering. (The distinction is the subject of our explainer on the neutrino mass ordering.)

A cosmological bound of 0.064 eV already sits almost on top of the normal-ordering floor. And when cosmologists apply a statistical correction that accounts for the fact that a physical mass cannot be negative — a Feldman–Cousins treatment — the 95 percent bound slides down toward 0.053 eV, beneath the oscillation floor. Taken at face value, the data prefer a neutrino lighter than oscillations allow. Some analyses that let the effective mass parameter run negative find a best fit below zero, which is not a real negative mass but a signal that cosmic structure is slightly more smoothed, or slightly less, than the simplest model with massive neutrinos predicts.

Is it the neutrinos, or the cosmology?

The honest answer is that nobody thinks the neutrino is breaking the rules of oscillation physics, which rest on decades of independent measurements. The pressure is far more likely to be on the cosmological model itself.

The tension largely evaporates when you loosen one assumption. If dark energy is allowed to evolve over cosmic time rather than being a fixed constant — the very possibility DESI’s own data have hinted at — the neutrino-mass bound relaxes to around 0.18 eV, comfortably above the oscillation floor. Swap in a different CMB dataset, such as SPT-3G, as the baseline, and the limit moves to roughly 0.11 eV. In other words, the “neutrino mass problem” may really be a message about dark energy wearing a neutrino disguise. This is now one of the liveliest debates at the border between particle physics and cosmology, and a recurring theme across our research and theory coverage.

What would settle it

Three things will sharpen the picture over the next few years. Cosmological surveys will keep accumulating data, tightening or shifting the bound and testing whether evolving dark energy survives. Laboratory experiments — KATRIN’s final campaign and the next generation aiming below 0.1 eV — will provide a model-independent cross-check that owes nothing to assumptions about dark energy. And JUNO, now taking data, should determine the mass ordering outright; if it confirms the normal ordering, it will agree with where cosmology is already pointing.

For now, the remarkable situation stands: two completely different ways of weighing the most elusive particle in nature have grown precise enough to disagree with each other. That disagreement is not an embarrassment. It is exactly the kind of crack through which new physics — about neutrinos, or about the dark energy filling the cosmos — tends to announce itself.

Frequently asked

What is the current cosmological bound on the neutrino mass?

Combining DESI Data Release 2 baryon acoustic oscillation measurements with cosmic microwave background data from Planck and ACT, cosmologists find the sum of the three neutrino masses to be below about 0.064 eV at 95% confidence in the standard ΛCDM model. This is several times tighter than any laboratory limit, including KATRIN.

Why is this in tension with oscillation experiments?

Oscillation experiments measure the differences between the squared neutrino masses, which sets a hard lower bound: the three masses must sum to at least about 0.059 eV for the normal ordering and about 0.10 eV for the inverted ordering. When a boundary-corrected version of the DESI bound dips toward or below 0.06 eV, it starts crowding — and in some analyses breaching — that oscillation floor, which should be physically impossible.

Does this mean neutrinos are weirder than we thought?

Most cosmologists think not. The more likely explanation is that the simplest cosmological model is too rigid. Allowing dark energy to evolve with time loosens the neutrino-mass bound to roughly 0.18 eV, and using a different CMB dataset as the baseline gives about 0.11 eV. The tension may be telling us something about dark energy rather than about neutrinos.

Has cosmology ruled out the inverted mass ordering?

Effectively, within standard ΛCDM. The inverted ordering needs Σm_ν above roughly 0.10 eV, well above the tightest cosmological bounds. That puts cosmology in agreement with the mild preference for the normal ordering seen in oscillation experiments like NOvA and T2K, and with early JUNO expectations.

Cite this article 5 formats

APA

Neutrino Times Editorial Team. (2026, June 22). Cosmology keeps shrinking the neutrino mass — and it is colliding with oscillations. Neutrino Times. https://neutrino-times.com/articles/cosmological-neutrino-mass-tension-desi/

Chicago

Neutrino Times Editorial Team. "Cosmology keeps shrinking the neutrino mass — and it is colliding with oscillations." Neutrino Times, June 22, 2026. https://neutrino-times.com/articles/cosmological-neutrino-mass-tension-desi/.

MLA

Neutrino Times Editorial Team. "Cosmology keeps shrinking the neutrino mass — and it is colliding with oscillations." Neutrino Times, 22 Jun. 2026, https://neutrino-times.com/articles/cosmological-neutrino-mass-tension-desi/.

BibTeX

@misc{neutrino-times-cosmological-neutrino-mass-tension-desi,
  author       = {Neutrino Times Editorial Team},
  title        = {Cosmology keeps shrinking the neutrino mass — and it is colliding with oscillations},
  howpublished = {Neutrino Times},
  year         = {2026},
  month        = {jun},
  url          = {https://neutrino-times.com/articles/cosmological-neutrino-mass-tension-desi/},
  note         = {Accessed: 2026-06-22}
}

RIS

TY  - GEN
TI  - Cosmology keeps shrinking the neutrino mass — and it is colliding with oscillations
AU  - Neutrino Times Editorial Team
PY  - 2026
DA  - 2026-06-22
PB  - Neutrino Times
UR  - https://neutrino-times.com/articles/cosmological-neutrino-mass-tension-desi/
ER  -