Photograph · 2MASS · CC BY-SA 3.0
Star · Deep guide
Stephenson 2-18
If it is what it seems, this may be the largest star known.
What is it?
Stephenson 2-18, an extreme red supergiant candidate ~19,000 light-years away, tops many 'largest stars' lists with a claimed radius near 2,150 Suns — big enough to swallow Saturn's orbit. Every part of that sentence carries caveats: its cluster membership, distance, and therefore size are all debated. It exemplifies both the scale of the universe and the humility good astronomy requires.
The deep dive
Researched for the Atlas from Wikipedia — Stephenson 2 DFK 1 (10,375 characters read) · updated Sep 20, 2026
01 First catalogued long before anyone knew what it was
Stephenson 2-18 entered the astronomical record quietly in 1988, logged under the dry designation IRAS 18363-0607 in the IRAS Point Source Catalog — essentially a warm infrared dot among millions. It attracted little attention until the open cluster around it came into focus. American astronomer Charles Bruce Stephenson discovered the cluster itself in 1990 using data from a deep infrared survey, and the cluster now carries his name. Even then, the star was identified simply as the brightest object in the region and given the identifier 1, only to be immediately set aside as probably unrelated to the cluster because of its outlying position, its abnormally high brightness, and a proper motion that didn't quite match the other stars. It wasn't until a 2010 study that it was renumbered 18 and tentatively placed in an outlying group called Stephenson 2 SW. The resulting tangle of competing designations — D1, DFK 1, RSGC2-01, St2-18 — reflects decades of researchers independently naming the same object, and conventions now recommend the DFK prefix to prevent confusion.
02 Why its distance is so hard to pin down Deeper
Measuring the distance to Stephenson 2-18 has proven stubbornly difficult, and the numbers have shifted dramatically over time. When Stephenson first announced the cluster in 1990, it was placed at roughly 30 kiloparsecs (about 98,000 light-years) — far enough to sit near the edge of the galaxy. By 2007, a kinematic analysis using radial velocity brought that estimate down to 5.83 kiloparsecs, with an asymmetric uncertainty of +1.91/−0.78 kiloparsecs, placing it around 19,000 light-years away. A 2010 study using the average radial velocity of four cluster members — 96 kilometers per second — derived a similar value of 5.5 kiloparsecs, but explicitly noted that the uncertainty in the distance exceeded 50%. That is an extraordinarily large margin: it means the star could be anywhere from roughly 9,000 to over 27,000 light-years away. Because almost every physical property — luminosity, radius, mass-loss rate — scales with assumed distance, that uncertainty cascades through every other measurement. The star's peculiar spectral energy distribution, which cannot be fitted to standard reddening laws, further complicates the picture by suggesting higher-than-assumed extinction.
03 The membership debate that changes everything
Whether Stephenson 2-18 actually belongs to its supposed cluster is not a settled question, and the answer matters enormously. If it is a genuine member of Stephenson 2 or its outlying group Stephenson 2 SW, then the distance to the cluster anchors the star's luminosity and size, making it one of the largest known stars. If it is merely a foreground red supergiant that happens to lie along the same line of sight, those record-breaking numbers evaporate. The key evidence is radial velocity. A 2013 study found that the cluster's average radial velocity is +109.3 ± 0.7 kilometers per second, while the star's own radial velocity is only 89 kilometers per second — a difference of 20 kilometers per second. Some researchers cite that gap as clear evidence of non-membership. Others argue the gap is actually quite ordinary: an expanding, optically thick stellar envelope could shift the measured radial velocity by exactly that amount, since 20 kilometers per second is a typical outflow speed for red supergiant winds. The debate remains open, and more recent papers have leaned toward considering the star an unlikely cluster member because of its extreme and internally inconsistent properties.
04 Luminosity estimates ranging across a factor of seven Deeper
Few stars have had their luminosity revised as dramatically or as repeatedly as Stephenson 2-18. The first published estimate, in 2010, used flux densities at 12 and 25 micrometers and returned a relatively modest 90,000 solar luminosities. The authors themselves noted this was likely an underestimate because it relied on a narrow range of fluxes. In 2012, the Australia Telescope Compact Array was brought to bear, fitting a full spectral energy distribution with the DUSTY radiative transfer model. That analysis gave a luminosity of nearly 440,000 solar luminosities at the same assumed distance of 5.5 kiloparsecs. The most recent calculation, assuming a distance of 5.8 kiloparsecs and integrating published fluxes across the full observed SED, pushes the value still higher to 630,000 solar luminosities. That is seven times the 2010 figure. Part of the spread is methodological — different flux ranges, different models — but part reflects the genuinely peculiar nature of the star's SED, which resists fitting to standard interstellar reddening laws. If the star suffers more extinction than standard models predict, its intrinsic luminosity would be higher still, and its distance — and therefore its cluster membership — becomes even harder to assess.
⤢
05 A radius that dwarfs our entire solar system
The 2012 DUSTY model analysis paired its luminosity estimate with an effective temperature of 3,200 Kelvin to derive a stellar radius of 2,150 solar radii — equivalent to about 1.50 × 10⁹ kilometers, or 10.0 astronomical units. To appreciate that figure, consider that Saturn orbits the Sun at a distance of 9.59 astronomical units. If Stephenson 2-18 were placed at the center of our solar system, its surface would extend beyond Saturn's orbit. The volume implied by that radius is around 10 billion times larger than the Sun's. These numbers push well beyond what stellar evolutionary theory comfortably allows: models predict that the largest, coolest, most luminous red supergiants should have radii of roughly 1,500 solar radii, temperatures around 3,500 Kelvin, and luminosities up to about 320,000 solar luminosities — the boundary set by the Hayashi limit and the Humphreys–Davidson limit. A star exceeding those limits would not be in hydrostatic equilibrium, meaning its own gravity and internal pressure are not balanced, and it would be expected to behave in unstable and dramatic ways.
06 Mass loss so extreme it may end the star's current life
Stephenson 2-18 is shedding mass at an estimated rate of roughly 1.35 × 10⁻⁵ solar masses per year — among the highest mass-loss rates measured for any red supergiant star. To put that in perspective, the Sun loses only about 2 × 10⁻¹⁴ solar masses per year through its solar wind, meaning this star is ejecting material at a rate roughly a billion times greater. The evidence for this violent outflow includes a significant infrared excess, which signals that large amounts of warm dust have formed in the expelled gas surrounding the star. The star also displays strong silicate emission, particularly at wavelengths of 10 and 18 micrometers — spectral signatures characteristic of dust-laden stellar winds. Researchers have noted that the star may have already undergone an extreme mass-loss episode and that it appears to be on the brink of ejecting its outer layers entirely. That process would likely transform it into a luminous blue variable or even a Wolf–Rayet star — radically different types of objects that represent later stages in the evolution of the most massive stars.
07 Masers as a probe of stellar winds Deeper
Among the most useful tools for studying the extended atmosphere and circumstellar envelope of Stephenson 2-18 are masers — naturally occurring microwave lasers produced when specific molecules are excited in the right conditions. In 2012, a study used the Australia Telescope Compact Array to observe maser emissions from Stephenson 2-18 and 56 other red supergiants across the galaxy. The star was observed again that same year in a dedicated study of maser types on red supergiant stars in clusters. Silicon oxide maser emission and infrared carbon monoxide absorption lines were used in a later study to assess the star's radial velocity and cluster membership. Water masers were also detected in the star's circumstellar envelope. Within Stephenson 2, only the stars with the highest bolometric luminosities show maser emission at all; Stephenson 2-18 and one companion star, designated D2, are the only cluster-region stars confirmed to have masers, indicating they have the highest mass-loss rates among the group. The presence of strong silicate emission features at 10 and 18 micrometers, combined with the maser detections, paints a picture of a star enshrouded in a thick, dusty, outflowing shell.
08 Spectral type tells a story of extreme coolness
Stephenson 2-18 has been assigned a spectral type between M5 and approximately M6, based on analysis of its CO-bandhead absorption features, its overall spectrum, and the characteristics of titanium oxide absorption lines. For context, the M spectral class is already the coolest category for stars visible to the naked eye, and a type of M5 or M6 is toward the extreme cool end even within that class. Ordinary red supergiants are rarely as late as M5. The star's effective temperature of 3,200 Kelvin, derived from the DUSTY model fitting, is consistent with its very late spectral type and is significantly cooler than the roughly 3,500 Kelvin temperature that evolutionary models predict as the coolest possible stable red supergiant. Several studies have specifically described Stephenson 2-18 as a "very late-type red supergiant." This extreme coolness, combined with its other anomalous properties — high luminosity, high mass-loss rate, large infrared excess — has led some researchers to suggest reclassifying it from a red supergiant to an extreme red hypergiant, similar in character to the well-studied star VY Canis Majoris.
⤢
09 Where it sits in the galaxy and why that matters
Stephenson 2-18 lies in the constellation Scutum, a small but rich region of sky when viewed in infrared light because it looks directly toward a dense section of the Milky Way's interior. The open cluster Stephenson 2, which the star is tentatively associated with, resides in the Scutum–Centaurus Arm — one of the galaxy's major spiral arms — at a distance of about 5.8 kiloparsecs (19,000 light-years) from Earth. This arm is notable for hosting several massive open clusters, each containing multiple red supergiants, making the region something of a hotbed for studying the most extreme evolved stars. The cluster itself blends into its immediate surroundings, and as one 2012 study noted, the stellar association is spread over a large area, making membership boundaries genuinely fuzzy. The Scutum region is so heavily obscured by interstellar dust that optical telescopes reveal little; the cluster was only discovered through deep infrared surveys, and nearly all study of stars in the region relies on infrared and radio wavelengths to penetrate the dust.
10 Instability and what it signals about the star's future Deeper
A star as cool, luminous, and large as Stephenson 2-18 appears to be — if its measured properties are taken at face value — should not, according to standard stellar physics, be in a stable state. The Hayashi limit defines the coolest temperature at which a star of a given mass can maintain hydrostatic equilibrium, and the Humphreys–Davidson limit describes an empirical upper boundary of luminosity and temperature beyond which massive stars are not found in stable configurations. Stephenson 2-18 appears to sit beyond both boundaries. Stars in such a position are expected to be dynamically unstable, undergoing violent and irregular mass-loss events rather than the steady, structured outflows of more typical red supergiants. The comparison objects cited in the literature — VX Sagittarii and Westerlund 1 W26 — are both stars known for spectral variability and high mass-loss rates. If Stephenson 2-18 is genuinely in this regime, it may be in a short-lived transitional phase of evolution, careening toward a dramatic transformation in its surface character before eventually exploding as a supernova or collapsing directly to a compact remnant.
11 Open questions that remain unresolved
Despite decades of observations, some of the most basic facts about Stephenson 2-18 remain genuinely uncertain. Its cluster membership is contested: three different studies have reached opposite conclusions based on the same radial velocity data, depending on whether they attribute the velocity discrepancy to non-membership or to an expanding circumstellar envelope. Its distance carries an acknowledged uncertainty exceeding 50%, which propagates into every derived quantity. Its spectral energy distribution cannot be matched to standard interstellar reddening laws, suggesting an unusual dust environment that current models do not cleanly handle. Its luminosity has been estimated at values ranging from 90,000 to 630,000 solar luminosities — a factor-of-seven spread. And its classification as either a red supergiant or an extreme red hypergiant has not been definitively settled. A 2020 paper by Humphreys and collaborators even mistakenly referred to the star by the name of a different large red supergiant in the same constellation, a small sign of how the field is still working to establish clear observational identities for these highly obscured, extreme objects.
⤢
