Stephenson 2-18 holds the larger size estimate of the two, with a frequently cited radius of roughly 2,150 times that of the Sun, compared to UY Scuti’s estimates that now range from about 750 to 1,700 solar radii depending on the study. But “which is bigger” turns out to be a surprisingly shaky question when applied to red supergiants, whose sizes are measured with error bars wide enough to swallow the difference between them. The real story is less about crowning a winner and more about why pinning down the size of any red supergiant is one of the hardest problems in stellar astrophysics.
What the Current Estimates Say
UY Scuti, located in the constellation Scutum roughly 5,000 light-years from Earth, was widely reported as the largest known star after a 2012 study estimated its radius at around 1,708 solar radii. That estimate relied on a particular distance measurement, and when later work revised the distance downward, the radius shrank with it. More recent analyses have placed UY Scuti closer to 1,000 or even 750 solar radii. That is still extraordinarily large, large enough that if placed at the center of our solar system it would engulf Jupiter’s orbit, but it is a far cry from the original headline figure.
Stephenson 2-18, also called St2-18, sits in the massive star cluster Stephenson 2, roughly 19,000 light-years away in the direction of the constellation Scutum. Its estimated radius of about 2,150 solar radii, if accurate, makes it potentially the largest star identified so far. At that size, its surface would extend past Saturn’s orbit. But Stephenson 2-18 is nearly four times farther away than UY Scuti, which compounds every source of measurement uncertainty. Its membership in the Stephenson 2 cluster helps constrain its distance somewhat, but the cluster distance itself carries meaningful error.
So on paper, Stephenson 2-18 wins. In practice, the uncertainties on both measurements overlap enough that the ranking could flip with the next round of observations.
Why Measuring a Red Supergiant Is So Difficult
Stellar radius is not measured with a ruler. For stars too distant to resolve directly, astronomers typically estimate radius from two things: how luminous the star is and how hot its surface is. A cooler star has to be physically larger to produce the same luminosity as a hotter one, so once you know the temperature and luminosity, you can back out the radius. The problem is that both luminosity and temperature are themselves derived quantities, and each depends on knowing the star’s distance accurately.
Distance is the weak link in the whole chain. The gold-standard method for stellar distances is parallax: measuring how much a star appears to shift against the background as Earth orbits the Sun. The European Space Agency’s Gaia mission has provided the best parallax data ever collected, but red supergiants present a unique headache. Their surfaces are not smooth, uniform disks. Convection churns the outer layers into enormous bright and dark patches, sometimes spanning a significant fraction of the stellar diameter. These patches cause the apparent center of the star’s light, its photo-center, to wander as the patches form and dissolve. Simulations of this effect show that the photo-center can shift by 1 to 5 percent of the star’s actual radius between successive measurements, introducing a systematic parallax error of a few percent even for relatively nearby red supergiants like Betelgeuse.1Astronomy & Astrophysics. Probing Red Supergiant dynamics through photo-center displacements measured by Gaia – Section: Results2Astronomy & Astrophysics. Photocentric variability of red supergiant stars and consequences on Gaia measurements
A few percent error in parallax may sound small, but distance enters the luminosity calculation as a square. If you underestimate the distance by 10 percent, you underestimate the luminosity by roughly 20 percent, and that ripples straight through to the radius estimate. For a star like UY Scuti, whose original distance estimate has been questioned, that kind of error is more than enough to account for the factor-of-two spread in published radii. For Stephenson 2-18, sitting four times farther away, the parallax signal is even tinier and the relative error even larger.
Where Does a Red Supergiant Actually End
Even if distance were perfectly known, there is a more fundamental problem: red supergiants do not have sharp edges. The Sun has a visible surface, the photosphere, that is well-defined because the transition from opaque to transparent happens over a very thin layer. In a red supergiant, that transition zone is enormously extended. The outer atmosphere can stretch outward for hundreds of solar radii, gradually fading from dense to tenuous, with no clean boundary.
The “radius” astronomers quote usually refers to the photospheric radius, the layer where most of the star’s light escapes. But which wavelength of light you observe changes where that layer appears. Infrared observations probe cooler, more extended layers than optical ones, so a star can look physically larger in infrared images. Molecular features like titanium oxide and vanadium oxide absorption, which are common in cool red supergiants, form in the extended atmosphere and can make the star appear bloated at certain wavelengths. WOH G64 in the Large Magellanic Cloud, long known as one of the most extreme red supergiants outside our galaxy, illustrates this problem: deep molecular bands and a possible detection of vanadium oxide indicate a highly extended atmosphere that complicates any single radius estimate.3Monthly Notices of the Royal Astronomical Society. A phoenix rises from the ashes: WOH G64 is still a red supergiant, for now
For both UY Scuti and Stephenson 2-18, the quoted radii assume a photospheric definition, but the exact boundary chosen, and the wavelength at which it is measured, varies between studies. Comparing two stars measured by different teams using slightly different methods and wavelengths introduces yet another layer of ambiguity to the “which is bigger” question.
Is There a Maximum Size for a Star
Red supergiants are not free to grow without limit. Two theoretical boundaries constrain how large and luminous they can become.
The first is the Hayashi limit. When a star’s envelope becomes fully convective, as happens in red supergiants, its structure settles onto a limiting configuration beyond which further radial expansion is not possible under hydrostatic equilibrium. Models show that stars with different core luminosities converge to a limiting effective temperature of around 4,000 Kelvin, corresponding to the nearly vertical track in the Hertzsprung-Russell diagram known as the Hayashi line.4arXiv. Why Do Stars Turn Red? – Section: 4.2 Physical Origin of the Red Giant/Supergiant Solution The region to the right of this line, meaning cooler and more expanded, is essentially forbidden because no stable hydrostatic solution exists there. This sets a ceiling on how bloated a red supergiant’s envelope can become at a given luminosity. A star that tries to expand past this limit would become dynamically unstable rather than remaining a well-defined, round object.
The second constraint is the Humphreys-Davidson limit, an observed upper boundary in the luminosity of red supergiants. Stars above roughly half a million times the Sun’s luminosity are not observed as cool supergiants; instead, they tend to shed mass violently and remain on the hotter side of the diagram. Recent modeling suggests that this boundary arises naturally from how much of a star’s weight is supported by radiation pressure rather than gas pressure. As a red supergiant approaches the point where radiation nearly supports the entire envelope, mass loss intensifies steeply, stripping material faster than the star can expand. This physically motivated mass-loss behavior reproduces the observed luminosity ceiling without requiring any artificial adjustments to the models.5Astronomy & Astrophysics. Exploring the Red Supergiant wind kink
Together, these limits mean that the very largest red supergiants, UY Scuti and Stephenson 2-18 among them, are living at the theoretical edge of what a star can be. They are pressing against both a maximum radius (Hayashi) and a maximum luminosity (Humphreys-Davidson). Any star that nominally exceeds both bounds simultaneously is probably being measured incorrectly, since stable objects cannot exist in that region of parameter space. This is worth keeping in mind when evaluating the most extreme published radius estimates for either star.
Other Stars in the Running
UY Scuti and Stephenson 2-18 get the most public attention, but the list of candidate “biggest stars” is longer and more contested than most popular accounts suggest.
WOH G64, located in the Large Magellanic Cloud, was for years considered the most extreme red supergiant outside the Milky Way. Its estimated radius has been placed as high as 1,500 to 1,800 solar radii in some analyses, though recent observations suggest it may be undergoing dramatic changes, possibly transitioning away from the red supergiant phase entirely.3Monthly Notices of the Royal Astronomical Society. A phoenix rises from the ashes: WOH G64 is still a red supergiant, for now Its highly extended atmosphere makes radius comparisons with Milky Way supergiants tricky, since the same atmospheric boundary issues apply.
Other Milky Way red supergiants with radii estimated above 1,000 solar radii include VY Canis Majoris, NML Cygni, and AH Scorpii. VY Canis Majoris was the reigning “largest star” in popular science for years before UY Scuti displaced it, and some recent distance revisions have actually pushed VY CMa’s estimated radius back upward. NML Cygni sits in a dense star-forming region that makes its distance and extinction particularly hard to pin down. The rankings among these stars shift every few years as new data arrive.
The common thread is that every star on the “biggest” list is plagued by the same measurement challenges: uncertain distances, extended and variable atmospheres, and wavelength-dependent apparent sizes. The title of “largest known star” has changed hands at least half a dozen times in the past two decades, and there is no reason to think it has settled.
Why the Biggest-Star Rankings Keep Changing
If you have looked up “the biggest star” more than once over the years, you have probably gotten a different answer each time. VY Canis Majoris was the answer through much of the 2000s, UY Scuti took the crown around 2012, and Stephenson 2-18 emerged as a contender more recently. This is not because astronomers are indecisive. It reflects genuine improvements in how distances and stellar properties are measured, combined with the intrinsic difficulty of the problem.
Each new data release from Gaia, for instance, updates the parallax measurements for thousands of stars simultaneously. A single revision to a star cluster’s distance can change the inferred luminosities and radii of every member star. The convection-driven photo-center wandering described earlier means that even Gaia’s precision is degraded for the exact class of stars people most want to measure.1Astronomy & Astrophysics. Probing Red Supergiant dynamics through photo-center displacements measured by Gaia – Section: Results Future Gaia data releases, which will include more epochs of observation and better modeling of extended sources, should help, but the systematic errors introduced by surface convection are unlikely to vanish entirely.
Interferometric observations, which combine light from multiple telescopes to resolve stellar disks directly, have provided some of the most reliable angular diameter measurements for closer red supergiants like Betelgeuse and Antares. But even interferometry struggles with more distant targets: at 19,000 light-years, Stephenson 2-18’s angular size on the sky is tiny, and resolving it directly is beyond current ground-based capabilities for most wavelengths. Until that changes, its radius will remain a derived quantity subject to all the compounding uncertainties already discussed.
There is also an underappreciated selection effect in the “biggest star” contest. The stars that make the list tend to be ones with the most uncertain measurements, because the measurement uncertainty itself is what allows an estimate to land in record-breaking territory. A star whose distance happens to be underestimated will appear overluminous and therefore oversized. The most extreme published radius for any given star is often an outlier rather than the consensus value, and popular accounts tend to pick up the extreme number because it makes for a better headline.
What Happens to These Stars Next
Both UY Scuti and Stephenson 2-18 are in the late stages of stellar evolution. Red supergiants of this size have already burned through their core hydrogen and helium and are fusing heavier elements in a series of nested shells. This phase is geologically brief: while the star may have spent millions of years on the main sequence burning hydrogen, the red supergiant phase lasts only a few hundred thousand years, and the final stages of heavy-element fusion can be measured in centuries or even decades.
The endpoint is a core-collapse supernova. When the core runs out of fusible material and collapses under its own gravity, the outer layers are blasted into space in an explosion visible across the universe. What remains is either a neutron star or, if the progenitor was massive enough, a black hole. The mass of the star at the moment of collapse, not its radius, determines which remnant forms. And that mass depends heavily on how much material the star lost during its red supergiant phase through stellar winds. The mass-loss rate itself is steep and sensitive to how close the star is to its theoretical luminosity limit, which is why modeling the Humphreys-Davidson boundary matters for predicting supernova outcomes.5Astronomy & Astrophysics. Exploring the Red Supergiant wind kink
Neither UY Scuti nor Stephenson 2-18 is expected to explode in our lifetimes, but on astronomical timescales, both are close to the end. When they do go, they will briefly outshine their entire host galaxies. Ironically, the explosion itself will be far easier to measure precisely than the radius of the star that produced it.
Pulsation and Variability Complicate Things Further
Red supergiants are not static objects. They pulsate, swelling and contracting over periods of hundreds to thousands of days. UY Scuti is classified as a semiregular variable, meaning its brightness and inferred size change noticeably over timescales of months. These pulsations are not subtle: the radius of a pulsating red supergiant can vary by 10 to 20 percent between maximum and minimum, which for a star in the thousand-solar-radii range translates to swings of a hundred solar radii or more.
On top of the pulsation, the convective surface structure introduces additional variability. The enormous bright granules on a red supergiant’s surface can be as large as the star’s own radius, and they evolve on timescales of months to years. As noted earlier, this convective motion shifts the apparent photo-center of the star by up to 5 percent of the radius, and the surface brightness pattern itself changes the effective temperature unevenly across the disk.2Astronomy & Astrophysics. Photocentric variability of red supergiant stars and consequences on Gaia measurements A radius measurement taken during one epoch may genuinely differ from a measurement taken a year later, not because of error, but because the star has physically changed.
This means that even if all systematic errors were eliminated, the question “which is bigger” would still lack a single fixed answer. Both stars are moving targets, literally expanding and contracting while their surfaces roil with convective cells larger than our entire solar system. Asking which is bigger at any given moment is a bit like asking which ocean wave is taller while they are both still cresting.