When more than a dozen ways of measuring the universe’s expansion all point to the same number, and that number still clashes with what the early cosmos predicts, the problem stops looking like a measurement mistake. An international team of astronomers reported in April 2026 that the universe is flying apart at roughly 73.5 kilometers per second per megaparsec, a figure now pinned down to about one-percent precision. The catch: the physics baked into the oldest light in the sky says the rate should be closer to 67.4. That six-unit gap, known as the Hubble tension, has persisted for more than a decade, and the latest result makes it harder than ever to explain away.
A measurement built on many rungs
The findings, published April 10 in the journal Astronomy & Astrophysics, come from a project called the Local Distance Network. Instead of staking everything on one cosmic yardstick, the collaboration combined more than a dozen distance indicators: parallaxes, Cepheid variable stars, Tip of the Red Giant Branch (TRGB) stars, Miras, JAGB stars, Type Ia supernovae, surface brightness fluctuations, the Fundamental Plane, and the Tully-Fisher relation. Weighting all of those together and accounting for their shared uncertainties produced the 73.5 km/s/Mpc value, with an uncertainty band tight enough to rule out most simple explanations for the discrepancy.
That number did not appear out of nowhere. It builds on a foundation laid by the SH0ES project, led in part by Nobel laureate Adam Riess of Johns Hopkins University. Using Hubble Space Telescope observations of Cepheid stars paired with Type Ia supernovae, SH0ES previously reported a local expansion rate of 73.04 plus or minus 1.04 km/s/Mpc. That measurement anchored the “late-universe” side of the tension and set the benchmark that every subsequent study has had to contend with.
A critical follow-up came from the James Webb Space Telescope. One of the most popular escape routes from the Hubble tension was the idea that Hubble’s cameras were blending Cepheid starlight with nearby stars in crowded galaxies, artificially inflating the measured rate. JWST’s sharper infrared vision can isolate individual Cepheids far more cleanly. When it did, it recovered essentially the same distances Hubble had found. Rather than resolving the tension, JWST deepened it by closing off one of the most plausible sources of error.
Separate calibration work reported in early 2026, using Gaia satellite parallaxes to refine TRGB and Magellanic Cloud distances, has pointed to a value broadly consistent with the Local Distance Network result. However, the specific figures from that analysis have not yet appeared in a peer-reviewed journal, so they should be treated as provisional. With Cepheids, TRGB stars, and multiple secondary indicators all converging on a similar expansion rate, the local side of the Hubble constant nonetheless rests on a broad and internally consistent foundation.
Why the gap still will not close
The other side of the tension comes from the cosmic microwave background (CMB), the faint afterglow of the Big Bang mapped in exquisite detail by the European Space Agency’s Planck satellite. When Planck’s data are interpreted through the standard cosmological model, known as Lambda-CDM, they predict an expansion rate near 67.4 km/s/Mpc. That prediction has its own impressive precision, and no updated Planck analysis responding directly to the 2026 Local Distance Network result has appeared yet. For now, the early-universe side of the debate rests on previously published Planck analyses and their internal consistency checks.
Methods that fall between the two camps have not broken the tie. A model-independent reconstruction of the expansion history, combining DESI DR2 baryon acoustic oscillation data, cosmic chronometers, and Pantheon+ supernovae through a statistical technique called Gaussian process regression, produced an intermediate estimate near 69.0 plus or minus 1.0 km/s/Mpc. That value splits the difference, which could mean the expansion rate changes over time in ways Lambda-CDM does not capture, or it could mean subtle systematic errors still hide in one or more datasets.
Strong-lensing time-delay cosmography, pursued by the TDCOSMO Collaboration, offers a conceptually different approach. By measuring how long light from a distant quasar takes to travel along different paths bent by a foreground galaxy’s gravity, astronomers can infer distances and, from those, the expansion rate. The method is elegant, but peer-reviewed studies have shown that selection biases and a mathematical ambiguity called the mass-sheet degeneracy can shift the inferred rate in ways that are difficult to pin down. Full covariance data from TDCOSMO’s integration into the broader distance network have not been publicly released, making it hard for outside groups to independently assess the method’s weight in the combined result.
A conceptual gap also remains at intermediate distances. No JWST survey targeting standard candles or gravitational-wave “standard sirens” at redshifts between about 0.5 and 2 has reported results yet. Programs are planned, but until they deliver, the mid-redshift regime, the stretch of cosmic history between the pristine early universe and the well-mapped local neighborhood, remains comparatively underconstrained.
Sorting strong evidence from provisional clues
Not all lines of evidence carry equal weight, and keeping them sorted helps make sense of the debate.
The strongest category consists of direct measurements published in peer-reviewed journals and confirmed by independent instruments. The SH0ES Cepheid-supernova ladder, validated by JWST, belongs here. So does the new Local Distance Network result, which folds in multiple distance indicators rather than depending on any single rung. These measurements are internally consistent, have survived extensive scrutiny, and any proposed solution to the Hubble tension must accommodate them.
A second category includes model-independent reconstructions like the DESI-based analysis. These studies deliberately avoid assuming a specific cosmological model, but they depend on the quality and cross-calibration of their input datasets. Their intermediate values are informative rather than definitive, and they hint that the expansion rate may not behave as a single clean number across all epochs.
A third category covers methods still wrestling with known systematics. Time-delay cosmography is powerful in principle because it measures expansion through an entirely different physical process: the bending of light around massive structures. But until selection-bias corrections and lens-modeling ambiguities are fully resolved, its constraints carry wider practical uncertainty than headline error bars suggest. For now, strong-lensing values serve as supporting evidence, not tiebreakers.
What the tension could be telling us
If the Hubble tension is real, and the accumulating evidence increasingly suggests it is, something in the standard model of cosmology needs updating. Theorists have floated a range of possibilities. One leading idea is “early dark energy,” a brief burst of accelerated expansion in the universe’s first few hundred thousand years that would have nudged the predicted expansion rate upward, closer to local measurements. Another involves neutrinos whose properties evolve over cosmic time, subtly altering the expansion history. More radical proposals tinker with general relativity itself at cosmological scales.
None of these ideas has won broad acceptance. Each solves part of the puzzle while creating new tensions with other observations. But the steadily improving precision of local measurements is forcing theorists to be increasingly specific about what new physics they invoke and what testable predictions it makes.
On the observational side, the next few years should bring sharper constraints from multiple directions: future Gaia data releases will refine the calibration of nearby distance indicators, dedicated JWST programs will push standard-candle measurements into the mid-redshift desert, and expanded DESI surveys will trace the expansion history with finer resolution. Each new dataset will either narrow the gap or confirm that it is a permanent feature of the cosmos.
For now, the scorecard is clear. The local expansion rate, measured with multiple tools and cross-checked by two flagship space telescopes, sits near 73.5 km/s/Mpc. The early-universe prediction, filtered through the standard cosmological model, remains near 67.4. The distance between those two numbers is small in absolute terms but enormous in its implications. It suggests the universe is expanding in a way that our best theoretical framework cannot fully explain, and as the error bars continue to shrink, the room for comfortable explanations is shrinking with them.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
