Highest-Energy Neutrino Ever Detected: What KM3NeT Found

The highest-energy neutrino reported so far is KM3-230213A, an event recorded by KM3NeT in the Mediterranean Sea on 13 February 2023. Its inferred neutrino energy is about 220 petaelectronvolts, with substantial uncertainty. A September 2026 IceCube follow-up found no supporting neutrino emission from its direction. That leaves the origin unresolved and constrains some proposed sources.

What made this neutrino unusual?

KM3NeT announced the event in February 2025, after analysing a remarkably bright signal in its ARCA detector. The discovery paper associated that signal with a cosmic neutrino far beyond the energies usually accessible to neutrino telescopes. The detection date and the publication date are different; this was not a new particle arriving in September 2026.

One PeV, or petaelectronvolt, is one million billion electronvolts. The estimate of 220 PeV corresponds to about 0.035 joule. That is an everyday-sized amount of energy concentrated in a single particle. The conversion is simply 220 × 10¹⁵ multiplied by 1.602 × 10⁻¹⁹ joule per electronvolt.

Energy alone does not reveal where a neutrino was produced. It gives physicists a demanding condition that any proposed source must meet: the process has to supply enough energy to the particles whose interactions ultimately make the neutrino. A spectacular energy estimate starts that investigation; it does not finish it.

What did KM3NeT actually detect?

The detector did not photograph a neutrino. According to KM3NeT’s account of the signal, its sensors recorded more than 28,000 photons over roughly two microseconds. The pattern was consistent with a single energetic muon crossing the detector. The collaboration estimated that muon’s energy at about 120 PeV.

A muon is an electrically charged particle. Moving through water faster than light travels in that medium, it produces Cherenkov light. This does not violate the speed limit in vacuum. The light’s arrival times and distribution let a telescope reconstruct a track. The detection principle is discussed in Halzen and Hooper’s neutrino-astronomy review.

A muon neutrino can make such a muon when it interacts with matter near the detector. Some energy goes into other products, and the muon can lose energy before entering the instrument. Consequently, the reconstructed muon energy and the inferred parent-neutrino energy answer different questions. Treating 120 PeV and 220 PeV as competing measurements would mix those questions together.

Why 220 PeV is an estimate, not an exact reading

The parent neutrino is reconstructed through simulations of neutrino interactions, muon propagation and the detector response. The inferred energy depends partly on the assumed distribution of incoming neutrino energies. The headline number should therefore travel with that assumption and its uncertainty.

In the Nature analysis, the median inferred energy is 220 PeV for an assumed incoming spectrum proportional to E⁻². Under that assumption, 68% of the simulated neutrinos compatible with the muon-energy estimate lie between 110 and 790 PeV. This is a simulation-derived interval, rather than a direct calorimeter reading of the neutrino.

The broad interval does not make the event uninteresting. It changes the precision of the claim: this is an exceptionally energetic neutrino candidate, while its exact energy remains uncertain. It also prevents an unjustified statement that an accelerator must produce precisely one narrow energy to explain the observation.

What did IceCube find in September 2026?

The IceCube follow-up paper, submitted on 10 September 2026, searched 15 years of data for neutrinos from the event’s direction. The analyses considered steady emission, flaring activity and searches connected to the KM3NeT detection time. None found significant evidence for a neutrino source in the searches performed.

The team set upper limits on point-source emission instead. Such a limit answers a conditional question: how much neutrino emission can be accommodated by the data under a specified source spectrum and analysis? It is not a declaration that absolutely no neutrinos came from that patch of sky.

The collaboration’s 23 September explanation also notes the central puzzle: the larger, longer-running IceCube experiment has not reported a neutrino at the same extraordinary energy. The new searches make some steady or transient point-source explanations more difficult, particularly when those models predict accompanying lower-energy neutrinos.

Does IceCube’s non-detection contradict KM3NeT?

The two results test different propositions. KM3NeT reconstructed one event in its own detector. IceCube looked for emission that a source model could produce in another instrument. Whether those propositions are compatible depends on the expected event rates, uncertainties and the assumed source population.

A detector’s physical size alone does not determine those rates. Its sensitivity depends on energy, arrival direction, the surrounding medium, analysis cuts and exposure. An ultra-high-energy neutrino can also be absorbed while crossing the Earth. The cross-experiment study compares the observation with other experiments’ null results. A valid comparison must use the response of each experiment, rather than multiplying a volume by the years it has operated.

The useful question is therefore whether a proposed source would have produced enough detectable IceCube events that their absence becomes unlikely. That can create genuine tension with a model without disproving the reconstruction of the KM3NeT event. Conversely, pointing out that the instruments differ does not automatically dissolve the tension.

What each stage of the analysis establishes

Observation	What it establishes
KM3NeT light pattern	Evidence for an exceptionally energetic through-going muon associated with a neutrino.
Neutrino-energy reconstruction	An energy distribution conditional on simulations and a spectrum assumption.
IceCube follow-up	No significant supporting emission in the tested searches; source-model constraints.
A possible object in the same direction	A candidate association requiring further statistical and physical tests.

Where could a neutrino this energetic come from?

One possibility is an astrophysical accelerator, such as an active galaxy with a powerful jet. Cosmic-ray particles accelerated there can interact with gas or radiation and produce neutrinos. Because neutrinos carry no electric charge, magnetic fields do not bend their trajectories as they bend those of charged cosmic rays. That makes them useful messengers, as our IceCube neutrino-astronomy explainer describes.

KM3NeT has investigated candidate blazars, active galaxies whose jets point approximately towards us. Its counterpart study compares observations at several wavelengths and considers possible associations. An object inside a directional uncertainty region is a candidate; proximity on a sky map is not sufficient evidence that it produced this neutrino.

Another possibility is a cosmogenic neutrino: an ultra-high-energy cosmic ray interacts with background radiation during its journey through space. The neutrino would then be produced along that journey, rather than directly inside the original accelerator. KM3NeT lists this among the origins under investigation. Its status is a hypothesis, not an established classification of KM3-230213A.

What evidence could settle the origin?

More events would change the problem. Repeated neutrinos from a consistent direction could support a source association. A population spread across the sky could instead help test a diffuse flux. Measurements over a range of energies would show whether the same source model describes both the exceptional event and less energetic neutrinos.

Observations of light, timing and cosmic rays can add independent constraints. They must be compared with the prediction of a specific mechanism: some environments absorb gamma rays, and some sources vary. A missing counterpart matters only in relation to what that mechanism predicts would be visible.

The unresolved source also has no bearing on whether neutrinos have mass; that question rests on separate oscillation evidence, explained in How We Know Neutrinos Have Mass. The present result is narrower and still remarkable: KM3NeT opened an observational window at an extreme energy, and IceCube has tightened the conditions that an explanation must satisfy.