What Is the Hubble Tension?

Background

In 1929, astronomer Edwin Hubble published observations showing that distant galaxies recede from us at speeds proportional to their distance, the first clear evidence that the universe is expanding. The rate of that expansion, now called the Hubble constant, has been measured and refined ever since, and for most of the twentieth century the main challenge was simply narrowing down its value with any precision at all.

By the 2010s, two largely independent methods had matured enough to measure the Hubble constant with real precision, and they stopped agreeing. The "distance ladder" method, most prominently developed by the SH0ES collaboration led by Adam Riess, calibrates distances using Cepheid variable stars and Type Ia supernovae in the relatively nearby, present-day universe, and consistently returns a value around 73 km/s/Mpc. The second method starts from the opposite end of cosmic history: the Planck satellite's precision measurements of the cosmic microwave background, the relic radiation from roughly 380,000 years after the Big Bang, combined with the standard cosmological model (ΛCDM) to project forward to today's expansion rate, consistently return a lower value around 67 km/s/Mpc. Neither team has found a flaw in the other's method despite years of mutual scrutiny, and the disagreement, now widely called the Hubble tension, has grown more statistically significant as both measurements have improved, standing at roughly 5 to 6 sigma by the mid-2020s, well past the 5-sigma threshold physics conventionally uses to rule out a chance fluctuation.

Main Theories

An undiscovered systematic error

The most conservative explanation holds that one or both measurements contain a subtle, undetected systematic error, some unaccounted-for bias in calibrating Cepheid distances, in modelling supernova brightness, or in interpreting the cosmic microwave background. Astronomers have spent over a decade actively hunting for exactly this kind of error, cross-checking Cepheid calibrations against independent distance indicators such as the tip of the red giant branch and, more recently, JWST observations, without resolving the discrepancy. The persistence and growing statistical significance of the gap, even as independent teams and instruments have joined the effort, has led most specialists to treat pure measurement error as an incomplete explanation, though not an impossible one, for the full size of the tension.

New physics beyond the standard cosmological model

A second line of explanation holds that the standard ΛCDM model used to project the CMB-based measurement forward to the present day is missing something, some additional physical ingredient in the early universe that would shift its predicted value upward and close the gap. Proposed candidates include "early dark energy," a form of dark energy that was briefly more influential shortly before the cosmic microwave background was emitted, then faded; modifications to how neutrinos or dark matter behaved in the early universe; and, more speculatively, a variation on the same evolving-dark-energy idea that DESI's 2024-25 results independently suggested may better fit large-scale structure data than a simple cosmological constant. Each proposal can be tuned to close some of the gap, but no single model has yet gained the community's broad support, and some proposed fixes create new tensions with other cosmological data even as they ease this one.

Unmodelled local structure

A more modest proposal holds that the local, present-day universe is not as smoothly uniform as the standard model assumes, and that the SH0ES team's distance-ladder measurement is picking up a genuine local effect, such as our galaxy sitting in an unusually large underdense region of space, rather than a universal expansion rate. A large enough local void could, in principle, make nearby space appear to expand faster than the cosmic average without requiring any new physics. This explanation remains actively debated: some large-scale galaxy surveys offer tentative support for an underdensity in our cosmic neighbourhood, while others find the required void larger and less probable than current structure-formation models comfortably allow.

Common Misconceptions

The Hubble tension is sometimes described in popular coverage as evidence that the Big Bang itself is wrong. It is not: the evidence for cosmic expansion and a hot, dense early universe, established independently by Hubble's original observation and reinforced by decades of subsequent astronomy, is not in dispute. The tension concerns a specific numerical parameter, how fast expansion is proceeding today, not the underlying framework, and even the most radical proposed resolutions modify a detail of the standard cosmological model rather than discard it.

It is also sometimes assumed that because both values are "close" (67 versus 73), the disagreement is minor. In precision cosmology the two figures are measured with error bars small enough that a roughly 9 percent difference represents a statistically serious, currently unexplained discrepancy rather than an acceptable margin of error.

Current Consensus

Cosmologists agree the Hubble tension is real, measured independently by multiple teams using different instruments and methods, and has grown more significant rather than less as precision has improved. What remains genuinely unresolved is its cause: whether an undiscovered systematic error, new physics in the early universe, unusual local structure, or some combination of these will ultimately explain it. Upcoming data from JWST, the Vera C. Rubin Observatory, and further DESI releases are expected to sharpen the picture considerably over the coming years, either by identifying a subtle flaw in one measurement or by strengthening the case for new physics.

Why This Mystery Endures

The Hubble tension endures as one of cosmology's most closely watched open problems because it is, unusually for the field, an argument between two well-established, independently cross-checked methods rather than between solid data and pure speculation. Both the distance-ladder and cosmic-microwave-background measurements have survived a decade of hostile scrutiny from rival teams motivated to find each other's mistake, and neither has broken. That is precisely what makes the standoff so interesting to physicists: a genuine anomaly of this kind, rather than a simple error waiting to be caught, is rare, and history offers a real precedent for what such anomalies can lead to. Dark energy itself was discovered because two supernova teams in 1998 refused to dismiss an unexpected measurement as error, and instead followed it to a Nobel Prize-winning revision of cosmology, a pattern many cosmologists hope, without yet being able to promise, the Hubble tension might repeat.

The stakes are also unusually clear. If the tension resolves toward new physics, it would mark the first confirmed crack in the standard cosmological model since dark energy and dark matter were themselves incorporated into it, potentially reshaping how the universe's contents and history are understood. If it resolves toward an overlooked systematic error instead, it becomes a case study in how subtle a mistake can be while still surviving years of professional scrutiny. Either outcome would be a genuinely significant result, which is why observatories across the world are actively racing to gather the data that might finally settle it. This page is part of this site's cosmology coverage, itself part of the wider scientific theories and frontiers cluster.