
- Key Takeaways
- A Disagreement That Survived Better Measurements
- What the Hubble Tension Actually Measures
- Building Cosmic Distances from Nearby Anchors
- Reading the Expansion Rate from Ancient Light
- How Much Room Remains for Measurement Errors?
- Why New Physics Is Difficult to Introduce
- Independent Tests Can Break Shared Assumptions
- What Would Count as a Convincing Resolution?
- Summary
- Appendix: Top Questions Answered in This Article
- Appendix: Glossary of Key Terms
Key Takeaways
- Different methods still produce conflicting estimates of the universe’s present expansion rate.
- Better measurements have narrowed several explanations without identifying a definitive cause.
- A successful solution must explain the discrepancy and preserve other well-tested cosmic observations.
A Disagreement That Survived Better Measurements
In April 2026, an international collaboration reported a present-day cosmic expansion rate of 73.50 kilometers per second per megaparsec, with an uncertainty of 0.81. The result came from a network of astronomical distance measurements rather than a single type of star. NASA’s account of the distance-network analysis described another substantial effort to investigate the Hubble tension, a disagreement that has persisted as observations have improved.
The difficulty appears when that measurement is compared with estimates inferred from the early universe. Under the standard cosmological model, observations of ancient radiation generally predict a lower present expansion rate. The European Space Agency’s Planck mission famously yielded approximately 67.4 kilometers per second per megaparsec in its 2018 analysis.
These are estimates of the same present-day quantity. The disagreement does not arise simply because the universe expanded differently billions of years ago. Cosmologists already include changing expansion in their calculations. The question is whether the history inferred from the early universe leads to the expansion rate measured through nearby astronomical objects.
A megaparsec equals approximately 3.26 million light-years. The unusual units describe how recession speed increases with distance in the smooth cosmic expansion. An expansion rate of 73.5 kilometers per second per megaparsec means that an additional megaparsec of separation corresponds, on average, to an additional 73.5 kilometers per second of recession.
That relationship describes large-scale expansion. Individual galaxies also move under the gravitational influence of their surroundings, and nearby gravitationally bound systems do not simply expand according to the cosmic average. Separating those local motions from the larger pattern is part of the measurement problem.
The controversy has substantial scientific consequences because the expansion rate helps connect cosmic distances, ages, composition, and structure formation. A correction to distance measurements would refine an established framework. Evidence for missing physics could require changes to the account of how the universe evolved.
Neither possibility should be declared the winner merely because it produces a more striking headline. The expansion-rate disagreement is important precisely because well-developed methods have become sufficiently precise to expose a mismatch. Understanding its cause requires examining the measurement chains and the assumptions connecting them.
What the Hubble Tension Actually Measures
The Hubble constant, usually written H0, describes the universe’s expansion rate at the present cosmic epoch. The word constant can mislead readers into assuming that expansion must have proceeded at that rate throughout history. In cosmology, the expansion rate changes as the relative contributions of radiation, matter, and dark energy change.
The broader quantity is the Hubble parameter, which can be evaluated at different times. H0 is its present value. Comparing measurements requires specifying whether an observation estimates today’s rate directly or measures an earlier stage of cosmic history from which today’s rate is inferred.
The standard framework is called Lambda cold dark matter, usually abbreviated ΛCDM. Lambda represents a cosmological constant associated with dark energy. Cold dark matter describes matter that does not emit light and moved slowly enough in the early universe to support the formation of cosmic structure.
Within this framework, the universe evolves according to general relativity, together with specified amounts of ordinary matter, dark matter, radiation, and dark energy. Its predictions include relationships among the expansion history, the distribution of galaxies, and the patterns in ancient radiation.
Understanding the distinction between dark matter and dark energy helps clarify the stakes. Dark matter contributes to gravitational clustering. Dark energy affects the expansion history. Adjusting either component can change predictions, but the changes must remain compatible with observations that constrain those components separately.
The Hubble tension is consequently a consistency test of an interconnected account of the universe. It compares observations made using different objects and physical processes, translated into a common parameter. Agreement would indicate that the observational methods and the connecting model fit together. Disagreement identifies something requiring further examination.
It is also important to distinguish precision from accuracy. Precision describes how tightly a measurement is constrained within its analysis. Accuracy concerns how close that estimate lies to the actual value. A highly precise result can still be biased if an unrecognized effect shifts the entire measurement.
Two narrow uncertainty intervals that do not overlap are significant, but they do not automatically establish new physics. The intervals describe uncertainties that researchers have modeled. Unknown or underestimated systematic effects can remain outside that accounting.
Conversely, invoking unspecified errors is not a sufficient explanation. A proposed measurement problem must be large enough, operate in the correct direction, and survive comparisons with independent data. The controversy remains productive only when both new-physics proposals and observational explanations face that same requirement.
Building Cosmic Distances from Nearby Anchors
One major approach measures distances outward from objects whose distances can be established geometrically. The resulting sequence is known as the cosmic distance ladder. Each step calibrates a method that reaches farther into space than the previous one.
For nearby stars, parallax provides a geometric starting point. As Earth changes position around the Sun, a nearby star appears to shift slightly against more distant objects. Measuring that angle allows astronomers to calculate distance without first assuming the star’s intrinsic brightness.
Other geometric anchors include carefully modeled binary stars and water masers orbiting galactic centers. A maser amplifies microwave radiation, and some galactic masers trace rotating disks whose motions and geometry support distance estimates. Multiple anchors reduce dependence on any single calibration route.
Cepheid variable stars provide a subsequent step. Their pulsation periods correlate with their intrinsic luminosities. Once that relationship is calibrated, astronomers can compare a Cepheid’s expected luminosity with its observed brightness to estimate its distance.
Cepheids can then establish distances to galaxies that have hosted Type Ia supernovae. These stellar explosions are bright enough to be observed much farther away. Their brightness is standardized using measured properties, allowing calibrated supernovae to connect the nearby distance scale to galaxies in the more uniform expansion.
Every connection introduces questions. Dust changes observed colors and brightness. Chemical composition can influence stellar behavior. Unresolved neighboring stars can contaminate a target’s light. Instrument calibration must remain consistent across observations made with different telescopes, filters, and exposure conditions.
The method is more demanding than identifying a familiar-looking star and assigning it a distance. Researchers must establish how the same class of object behaves in different environments and demonstrate that their corrections transfer reliably between nearby calibrators and distant targets.
The Local Distance Network explicitly addresses the fact that distance methods can share calibrations and data. Its baseline result combined selected indicators with covariance weighting, which accounts for relationships among their uncertainties. The analysis also tested the consequences of removing particular indicators or replacing parts of the network.
That approach illustrates an important principle: counting methods is not equivalent to counting independent experiments. Several apparently separate estimates can inherit the same calibration uncertainty. A reliable combination must identify shared information rather than allowing it to create an artificial impression of certainty.
The distance ladder has become increasingly sophisticated because its vulnerabilities are measurable. Improvements arise from testing those vulnerabilities individually and from constructing alternative routes through the distance network.
Reading the Expansion Rate from Ancient Light
The second major approach begins with the cosmic microwave background, or CMB. This radiation was released when the early universe cooled enough for neutral atoms to form and light could travel relatively freely. It provides a record of conditions approximately 380,000 years after the beginning of the hot Big Bang expansion.
The cosmic microwave background is not a direct photograph of galaxies or stars. It records tiny variations associated with the distribution and motion of matter and radiation before the first galaxies formed.
Before the radiation was released, pressure and gravity produced oscillations in the hot mixture of ordinary matter and light. These processes left characteristic patterns in the CMB. Their angular sizes and relative strengths encode information about the universe’s contents and geometry.
A particularly important scale is the sound horizon: the distance sound waves could travel through the early cosmic plasma during the relevant period. The observed angular size of that scale helps constrain the relationship between physical distances in the early universe and the expansion history leading to the present.
Inferring H0 requires a model to connect those stages. Researchers fit the observed radiation patterns with predictions, estimate cosmological parameters, and calculate the present expansion rate implied by the resulting history. The Planck-based estimate is a model-dependent inference from very precise observations.
Model dependence does not make the estimate arbitrary. The same framework must explain multiple aspects of the radiation pattern simultaneously. Changing one ingredient can improve a particular comparison and worsen others, which gives the analysis considerable power to reject unsuitable alternatives.
However, the distinction between observation and inference remains essential. Planck measured the sky’s microwave radiation. It did not observe today’s expansion rate in the same manner as a calibrated nearby distance survey. If the physical model connecting early and late times is incomplete, the inferred H0 could be shifted even when the radiation measurements are accurate.
This creates a specific path toward new physics. A change to conditions before the CMB formed could alter the calculated sound horizon. The observed angular pattern might then be consistent with a different expansion history and a higher present value of H0.
The difficulty is that the sound horizon does not operate independently of the rest of cosmology. A proposed change also affects other features of the CMB, the formation of structure, or the abundance of light elements. A convincing solution must satisfy the combined evidence rather than merely move one number.
How Much Room Remains for Measurement Errors?
Systematic errors differ from random measurement noise. Random fluctuations can diminish as observations accumulate. A systematic effect can persist across a larger sample, shifting many measurements in the same direction and making an incorrect answer appear increasingly precise.
Stellar crowding is a concrete example. A distant Cepheid may be observed together with unresolved neighboring stars. The extra light can make it appear brighter than it truly is. If the contamination is underestimated, the inferred distance can become too small.
The James Webb Space Telescope provides a useful test because its infrared observations can separate stars more clearly in relevant circumstances. NASA reported in March 2024 that Webb and Hubble comparisons supported the Cepheid measurements across the distances examined. That finding weakened a specific explanation in which increasing crowding errors would account for the tension.
The scope of that result matters. Demonstrating that one proposed bias is insufficient does not establish that every possible source of error has vanished. It strengthens a particular part of the measurement chain and narrows the remaining explanations.
Alternative stellar indicators provide another check. The tip of the red giant branch uses a recognizable brightness boundary associated with a stage in stellar evolution. Carbon-rich asymptotic giant branch stars offer another potential distance indicator. These populations differ from Cepheids, so their environmental sensitivities and measurement problems are not identical.
The Chicago-Carnegie Hubble Program examined such comparisons using Webb observations. Its work illustrates why disagreements among distance indicators deserve attention even when they do not immediately resolve the broader cosmological problem. Small shifts in calibration can matter when the intended precision approaches the percent level.
Supernova measurements introduce additional considerations. Surveys must account for which explosions are detected, how brightness corrections are applied, and whether nearby calibrating events differ systematically from more distant samples. A faint event is harder to detect than a bright one, so the observed sample is not automatically representative.
Local motions also complicate nearby measurements. Galaxies fall toward concentrations of matter and move away from underdense regions relative to the average expansion. Researchers attempt to model or reduce these effects by selecting appropriate distance ranges and using information about surrounding structure.
The scientific standard is an identified mechanism with demonstrable consequences. Neither a general assurance that instruments are accurate nor a general suspicion that astronomy is difficult settles the issue. Each proposed bias must be tested quantitatively against the actual observations.
Why New Physics Is Difficult to Introduce
The persistence of the discrepancy has motivated theories that modify the early universe, the later expansion history, or gravity itself. These proposals differ substantially, and the phrase new physics should not be treated as the name of a single competing explanation.
Early dark energy is one prominent category. Such models introduce a temporary contribution to the cosmic energy budget before or around the era relevant to the formation of the CMB. By changing the early expansion rate, they can reduce the sound horizon and shift the inferred present expansion rate.
The term can be confused with proposals that dark energy changes during comparatively recent cosmic history. The two ideas concern different epochs and do not automatically solve the same observational problems. A model designed to address late-time expansion measurements may leave the Hubble tension largely unchanged.
A review of early dark energy describes both the appeal of the mechanism and the difficulty of satisfying the broader evidence. The challenge is not merely finding a mathematical adjustment that raises H0. It is identifying a physically credible adjustment that remains consistent with other observations.
Additional light particles offer another route. Extra radiation-like energy in the early universe would influence its expansion and the acoustic scales used in cosmological inference. However, the radiation content is also constrained by other measurements, including the production of light elements and the detailed CMB pattern.
Modified-gravity proposals change the relationship between matter, spacetime, and cosmic evolution. They face a particularly demanding task because gravity has been tested in many environments. A modification relevant on cosmological scales must explain why successful predictions remain intact in the settings where those tests apply.
Changes to the local cosmic environment have also been examined. A sufficiently unusual distribution of nearby matter could affect locally inferred expansion. Such explanations require evidence that the relevant structure exists and is large enough to produce the observed shift without conflicting with other surveys.
A useful way to assess these proposals is to ask what additional observation each predicts. A theory becomes more informative when it specifies a measurable consequence beyond the discrepancy it was designed to address. Otherwise, added flexibility can absorb a mismatch without explaining its origin.
The age of the universe supplies another consistency requirement. Changing expansion history can change the inferred cosmic age. That age must remain compatible with independently studied old stellar populations, allowing for uncertainties in stellar evolution and formation times.
Independent Tests Can Break Shared Assumptions
The strongest route toward resolving the controversy is a collection of measurements that do not all depend on the same calibration chain. Independence is valuable because an error shared by several methods can survive repeated observations with impressive internal agreement.
Gravitational-wave standard sirens provide one such approach. The waveform from a compact-object merger contains information about distance. If the source’s host galaxy can be identified, its redshift supplies the corresponding information about cosmic expansion.
The gravitational-wave measurement method bypasses the need to calibrate Cepheid brightness. It introduces different uncertainties, including the source’s orientation, detector sensitivity, host identification, and assumptions used for events without a clearly identified electromagnetic counterpart.
That difference in vulnerabilities is the method’s scientific value. It does not make the measurement automatically more accurate than a distance ladder. It creates an opportunity to test whether an independently constructed result converges on the same expansion rate.
Other approaches use the timing of multiple images produced when a massive foreground object gravitationally lenses a variable background source. Light follows different paths to form the images, so changes in the source arrive at different times. Interpreting those delays requires a sufficiently accurate model of the lensing mass.
Distances derived from masers, alternative stellar indicators, and the apparent structure of galaxies provide further checks. Each method trades one set of observational requirements for another. The goal is a network whose conclusions remain stable when a particular indicator or calibration is removed.
The earliest stages of cosmic history also offer independent information. Light-element abundances constrain conditions during primordial nucleosynthesis. Combined with suitable large-scale-structure measurements, they can test aspects of expansion without relying on precisely the same CMB inference route.
Stellar ages contribute differently. Studies of the oldest Galactic stars do not directly measure H0, but they can constrain cosmological histories that would make the universe younger than its oldest surviving populations. Their uncertainties must include both stellar modeling and the interval before those stars formed.
A successful resolution may emerge from an unexpected direction. A distance indicator could reveal a calibration problem, a gravitational-wave sample could favor one scale, or early-universe observations could rule out a proposed physical modification. The relevant advance is the elimination of shared assumptions, not simply the accumulation of another headline value.
What Would Count as a Convincing Resolution?
A measurement-based resolution would identify a specific effect, demonstrate its magnitude, and show that correcting it brings previously conflicting results into agreement. The correction would need to work across the relevant observations rather than being chosen solely because it produces the desired answer.
For example, a revised calibration would become persuasive if it were supported by independent geometric distances and consistently changed multiple downstream indicators. An unexplained offset inserted into an analysis would have much less explanatory value.
A new-physics resolution would face a broader burden. It would need to improve the agreement among expansion measurements, fit the established cosmic observations, and preferably predict an additional signal that could be tested independently. The evidence would be stronger if different teams recovered that signal using different instruments and analysis methods.
Statistical significance is part of this evaluation, but it cannot substitute for understanding the measurement. A discrepancy expressed in standard deviations depends on the adopted uncertainty model, correlations, and comparison. Different data combinations can legitimately yield different significance estimates.
The same caution applies to claims of consensus. A collaboration can reach a carefully defined consensus measurement without every researcher agreeing that a particular physical interpretation follows. Agreement about the value obtained from a specified network is narrower than agreement about the universe’s underlying explanation.
The most useful reporting separates three questions: whether a discrepancy appears in a particular comparison, whether known observational effects can explain it, and whether a specific alternative theory has been established. Progress on the first two does not automatically answer the third.
There is also no requirement that the final explanation fit neatly into either measurement error or revolutionary physics. Several smaller corrections could combine with a modest extension of the standard model. Alternatively, a more complete treatment of existing physical processes could alter an inference without introducing a new fundamental ingredient.
Cosmology’s strength is that its claims are connected. The same universe must account for ancient radiation, nearby stellar distances, galaxy clustering, light-element abundances, and stellar ages. This interconnectedness makes the Hubble tension difficult to solve, but it also prevents an explanation from succeeding through an isolated numerical adjustment.
The unresolved status of the problem should be treated as a precise description of the evidence. The disagreement is substantial enough to justify continued investigation. The observations do not yet establish a unique cause, and the most defensible account preserves both facts.
Summary
The Hubble tension concerns incompatible estimates of the universe’s present expansion rate obtained through different observational and theoretical routes. Nearby distance measurements build outward from calibrated astronomical objects. Early-universe estimates use ancient radiation and a cosmological model to infer what today’s expansion should be.
Improved observations have tested important weaknesses in the distance ladder, including stellar crowding. Broader distance networks have also reduced reliance on a single indicator. These developments make the disagreement harder to dismiss, but they do not establish that every systematic uncertainty has been eliminated.
Proposals involving early dark energy, additional particles, altered gravity, or the local cosmic environment must satisfy more than the expansion-rate comparison. They must preserve the successful explanation of other observations and offer testable consequences.
The most informative next steps are measurements with different assumptions. A convincing resolution will explain why the estimates disagreed and demonstrate that the explanation works across independent evidence. Until then, the Hubble tension remains a serious test of cosmology rather than a completed discovery of its replacement.
Appendix: Top Questions Answered in This Article
What Is the Hubble Tension?
The Hubble tension is a disagreement between estimates of the universe’s present expansion rate. Some measurements calibrated through nearby astronomical objects produce a higher value than estimates inferred from early-universe observations using the standard cosmological model. The comparison concerns today’s expansion, not simply different rates at different times.
Does the Hubble Constant Change Over Time?
The Hubble constant refers specifically to the present value of the cosmic expansion rate. The broader Hubble parameter changes as the universe evolves and its energy components change in relative importance. Calling the present value a constant does not imply that expansion has always proceeded at the same rate.
What Does a Megaparsec Mean?
A megaparsec is a distance of approximately 3.26 million light-years. In the units used for the Hubble constant, it expresses how much additional recession speed corresponds to additional separation in the smooth cosmic expansion. Individual galaxies can also have motions caused by nearby gravitational structures.
Why Are Cepheid Stars Useful?
Cepheids vary in brightness with periods related to their intrinsic luminosities. Once that relationship is calibrated using independently measured distances, their observed brightness can help establish distances to other galaxies. Their usefulness depends on careful treatment of dust, chemical composition, neighboring stars, and instrument calibration.
Does Webb Prove That Measurement Errors Are Impossible?
Webb observations have tested particular potential errors, particularly contamination from neighboring stars in Cepheid measurements. Supporting those measurements weakens specific explanations for the discrepancy. It does not logically eliminate every systematic uncertainty elsewhere in the distance network or establish a unique new-physics explanation.
Why Does Planck Need a Cosmological Model?
Planck measured ancient microwave radiation rather than the present expansion directly. Researchers use a model to connect the conditions recorded in that radiation with the universe’s later evolution. The inferred expansion rate depends on both the observations and the physical assumptions connecting early and late times.
Could Early Dark Energy Solve the Problem?
Early dark energy could change the expansion before the cosmic microwave background formed and reduce the calculated sound horizon. That can shift the inferred present expansion rate. Any successful version must also fit the detailed radiation pattern, structure formation, and other observations that constrain early cosmic conditions.
Can Gravitational Waves Help?
Gravitational waves from merging compact objects contain information about distance. Combined with suitable redshift information, they provide an expansion measurement that does not require Cepheid calibration. Their uncertainties differ from those of stellar distance methods, making them valuable independent tests rather than automatically decisive measurements.
Does the Tension Mean the Big Bang Is Wrong?
The discrepancy does not by itself overturn the evidence for an expanding universe that was hotter and denser in the past. It tests particular measurements and the detailed model connecting cosmic epochs. A resolution could involve calibration changes, additional physics, or a combination of effects.
What Would Settle the Disagreement?
A convincing resolution would identify the cause of the mismatch and reproduce consistent results across independent observations. A measurement explanation would need a demonstrated bias with the correct size and direction. A new-physics explanation would also need to preserve successful cosmological predictions and pass additional tests.
Appendix: Glossary of Key Terms
Hubble Constant
The present-day rate of cosmic expansion, conventionally expressed in kilometers per second per megaparsec. It relates average recession speed to distance on large scales and is distinct from the expansion rate at earlier periods in cosmic history.
Megaparsec
A unit of astronomical distance equal to one million parsecs, or approximately 3.26 million light-years. Cosmologists use it to describe separations between galaxies and the relationship between distance and the universe’s large-scale expansion.
Lambda Cold Dark Matter
The standard cosmological framework combining ordinary matter, cold dark matter, radiation, and a cosmological constant. It uses general relativity to connect the universe’s expansion history with observations of ancient radiation, galaxies, and the growth of structure.
Cosmic Distance Ladder
A connected sequence of methods for measuring astronomical distances. Nearby geometric measurements calibrate objects observable farther away, which then calibrate additional indicators. Uncertainties and shared calibrations must be carried through the sequence to the final distance estimate.
Cosmic Microwave Background
Ancient radiation released when the universe became sufficiently transparent for light to travel relatively freely. Its small temperature and polarization variations preserve information about early cosmic conditions and help constrain cosmological models and their predicted expansion histories.
Systematic Error
A measurement or modeling effect that shifts results in a consistent way rather than producing purely random scatter. Increasing the number of observations does not necessarily remove it, so researchers must identify and test the physical or analytical mechanism involved.
Standard Siren
A gravitational-wave source whose signal provides information about its distance. When combined with redshift information, it can constrain cosmic expansion without relying on stellar brightness calibration, although detector, source, and population uncertainties still require careful treatment.

