The radius of the event horizon of Sagittarius A* is roughly 12.7 million kilometres, about 18 times the radius of the Sun. That figure applies to the Schwarzschild radius of a non-rotating black hole of 4.3 million solar masses, which is the standard mass estimate for the Milky Way's central black hole from the GRAVITY Collaboration and the Event Horizon Telescope. A spinning black hole's actual horizon sits slightly inside that radius, so the number is an upper bound set by mass alone.
The keyword asks for one number, and the number depends entirely on one input: the mass. Nail the mass down and the radius follows from a formula that takes thirty seconds to evaluate. Get the mass wrong, or mix up the Schwarzschild radius with the photon sphere or the shadow the Event Horizon Telescope actually photographed, and you land on a figure that's off by a factor of 1.5 to 2.6.
I evaluated the formula directly from the published mass figure, so everything below is arithmetic you can redo, not a claim about what I looked up. The article walks through the formula, the mass estimate and its uncertainty, how the answer shifts under rotation, why the famous EHT image is much larger than the horizon itself, and how the Milky Way's black hole compares with M87* and with human-scale distances.
The Schwarzschild Formula That Converts Mass to Horizon Radius
The calculation rests on a single expression: the Schwarzschild radius equals 2GM/c², where G is the gravitational constant, M is the black hole's mass, and c is the speed of light. Physicists compress this into a handy shortcut: r_s in kilometres equals 2.953 times the mass in solar masses. That constant absorbs G, c, and the Sun's mass, so you only need the mass in suns.
Here's the arithmetic for Sagittarius A*. Take 4.3 million solar masses and multiply by 2.953 km per solar mass. The product is roughly 12.7 million kilometres. Run it with 4.0 million and you get about 11.8 million km; run it with 4.6 million and you get about 13.6 million km. The spread between those two extremes of roughly 1.8 million km is wider than the radius of the Sun, which tells you the mass uncertainty matters more than most popular accounts admit.
Why does a shortcut exist at all? Because 2GM/c² is linear in mass. Double the mass and the horizon doubles; there's no curve, no threshold, no hidden nonlinearity. That linearity is what makes the formula useful for quick comparison across objects — you can scale any black hole's radius from any other's without redoing the full calculation, as long as you know the mass ratio.
One caution about the constant. The value 2.953 km per solar mass uses the modern IAU-nominal gravitational constant and solar mass parameter. Older textbooks sometimes print 2.95 or 3.0 depending on the constants in vogue when they were written. The difference is under 0.2 percent, negligible for Sagittarius A*, but it explains why two sources can quote 12.7 versus 12.6 million km without either being wrong.
Why the Mass Estimate Anchors the Whole Answer
The 12.7 million km figure is only as good as the 4.3 million solar mass input, and that input comes from two independent measurement programmes that agree remarkably well. The GRAVITY instrument on the Very Large Telescope Interferometer tracked stars orbiting the galactic centre and, in 2018 and 2020 papers, pinned the mass at about 4.297 million solar masses with a formal uncertainty near 0.1 percent. The Event Horizon Telescope's 2022 imaging result independently reported roughly 4.0 million solar masses with wider error bars.
The GRAVITY number is the tighter constraint, and it's the one I'd use for the headline figure because stellar orbits give a direct dynamical mass. The stars are visible, their positions are measured to microarcsecond precision, and Kepler's laws do the rest. The EHT mass comes from fitting a ring diameter to a theoretical model, which introduces more modelling dependence. When two methods disagree by 7 percent, pick the one with the smaller systematic error and say so.
The 2020 GRAVITY paper also reported a distance to the galactic centre of about 8,247 parsecs, roughly 26,900 light-years. That matters because the mass measurement in solar masses depends on the distance: the orbital velocities are observed, and converting them to a mass requires knowing how far away the stars are. A 1 percent distance error propagates to roughly a 1 percent mass error, which becomes a 1 percent radius error. The distance and mass uncertainties are entangled.
Here's the part most articles skip. The mass estimate carries a stated uncertainty, and the radius inherits it directly. Quote the radius as 12.7 million km and you're implicitly claiming a precision the underlying data doesn't support beyond roughly plus or minus 0.3 million km. A more honest phrasing is "about 12.7 million km, with an uncertainty of a few hundred thousand kilometres from the mass measurement alone." The uncertainty is not glossed over as negligible.
How Spin Shrinks the Horizon Below the Schwarzschild Value
Real black holes rotate, and rotation changes the horizon size. For a Kerr black hole with spin parameter a, the outer horizon radius is r_+ = GM/c² + sqrt((GM/c²)² − (a/c)²), which reduces to the Schwarzschild value only when spin is zero. At maximum spin (a = GM/c), the horizon shrinks to exactly half the Schwarzschild radius. That's the theoretical floor for a given mass.
What is Sagittarius A*'s spin? The honest answer is that it isn't firmly measured. Estimates from X-ray reflection spectroscopy and from the EHT's ring morphology suggest moderate to high spin, with values quoted between roughly 0.5 and 0.9 of the maximum, but the error bars are large and the methods disagree. Genzel and colleagues' 2003 work and later X-ray studies point toward fast rotation; some modelling favours lower values. No published measurement pins it down the way the mass is pinned.
Run the numbers anyway, because the range is instructive. At spin 0.5, the horizon shrinks from 12.7 to about 11.3 million km. At spin 0.9, it drops to roughly 8.6 million km. At the maximum of 1.0, it's 6.35 million km. So the true horizon could plausibly sit anywhere from about 8.6 to 12.7 million km depending on spin, and that entire range is consistent with current observations. The 12.7 figure is therefore an upper bound, not a precise measurement.
This is where the answer's precision breaks down. If you want a single number for a school answer, 12.7 million km is defensible because it uses only the well-measured mass. If you want the physical horizon of the actual object, you have to admit you don't know it better than the spin range allows. State the Schwarzschild value, then state the caveat; don't present the caveat-free number as if it were the measured radius.
The 2019 and 2022 Images Show a Ring Far Larger Than the Horizon
The Event Horizon Telescope image released in 2022 shows a bright ring around a dark central region, and the ring's diameter is about 51.8 microarcseconds. At the 26,900-light-year distance, that translates to a ring diameter of roughly 6.6 billion kilometres, or a radius near 3.3 billion km. Compare that with the 12.7 million km horizon radius and you see the ring is about 260 times larger.
Why the enormous gap? The bright ring is not the horizon. It's the photon ring, formed by light that has orbited the black hole before escaping, and its apparent size is set by the photon sphere at 1.5 times the Schwarzschild radius, then gravitationally lensed outward. The shadow inside the ring is larger than the horizon too, roughly 2.6 times the Schwarzschild radius in apparent size for a non-spinning hole. The EHT measures the shadow and the ring, not the horizon itself.
This creates the single most common confusion in popular coverage. Headlines said the telescope "photographed the black hole," but what it imaged was the silhouette and surrounding emission, structures that sit well outside the event horizon. The horizon is invisible by definition — no light escapes it — so no telescope can image it directly. The 12.7 million km radius is an inferred quantity, not something in the picture.
The same scaling applies to M87*, the galaxy M87's black hole, which the EHT imaged in 2019. M87* has a mass of about 6.5 billion solar masses, roughly 1,500 times Sagittarius A*'s mass, so its Schwarzschild radius is about 19 billion km. Its ring appears larger on the sky despite being 55 million light-years away because the object itself is so much bigger. The comparison is useful for checking you've understood the scaling: same physics, different mass, different radius.
Sagittarius A* Versus M87*: A Direct Comparison Table
| Property | Sagittarius A* | M87* |
|---|---|---|
| Mass (solar masses) | 4.3 million | 6.5 billion |
| Schwarzschild radius | 12.7 million km | 19.2 billion km |
| Distance from Earth | 26,900 light-years | 55 million light-years |
| EHT ring diameter | 51.8 microarcseconds | 42 microarcseconds |
| Horizon radius in AU | 0.085 | 128 |
| Year first imaged | 2022 | 2019 |
The table makes the point that apparent size and physical size are different things. M87*'s ring looks smaller on the sky than Sagittarius A*'s despite being 1,500 times larger in physical radius, because it's 2,000 times farther away. The two effects nearly cancel, which is why both were just barely resolvable by the same telescope array.
Notice the horizon radius in astronomical units. Sagittarius A*'s horizon would fit comfortably inside Mercury's orbit, which sits at about 0.39 AU. M87*'s horizon at 128 AU would swallow the entire inner solar system out past Pluto, which orbits at roughly 39 AU. That comparison lands harder than raw kilometre figures for most readers, and it's derived directly from the mass ratio.
The distance column also explains a measurement subtlety. Because Sagittarius A* is close, its mass can be measured dynamically from stellar orbits. Because M87* is far, its mass comes from gas kinematics and from the EHT ring itself, which is why M87*'s mass uncertainty is larger in relative terms. The closer object gets the better measurement, and that's why the Milky Way's black hole gives the cleaner radius figure.
Putting 12.7 Million Kilometres Into Human Terms
Raw kilometre counts don't mean much to most people, so anchor the number. Light travels 12.7 million km in about 42 seconds. If you could shine a laser at the horizon and time the reflection, the round trip would take roughly 85 seconds. That's the light-crossing time of the horizon radius, and it's a useful sanity check: the horizon is small on cosmic scales but large on human ones.
Distance comparisons help more. The Moon orbits at about 384,000 km, so the horizon radius is about 33 times the Earth-Moon distance. The Sun's radius is about 696,000 km, so the horizon is about 18 solar radii. If you placed the black hole where the Sun is, the horizon would extend to roughly 18 times the Sun's surface radius, well past Mercury's orbit at 0.39 AU (58 million km) — actually inside it, at 0.085 AU.
Travel time gives another handle. The fastest human-made object, the Parker Solar Probe, reached about 635,000 km/h at its closest solar approach. At that speed, crossing one horizon radius would take about 20 hours. Crossing the diameter, twice that, about 40 hours. A crewed spacecraft at Apollo-era speeds of roughly 39,000 km/h would need about 13 days to cover the radius.
These comparisons aren't just decoration. They let you check whether a quoted figure is plausible. If someone says the horizon is "the size of a solar system," ask which one: 12.7 million km fits inside Mercury's orbit but is far smaller than the orbit of Neptune. The correct scale is "a small fraction of an astronomical unit," not "a solar system" writ large.
Where the Number Goes Wrong: Common Errors and Edge Cases
The most frequent error is confusing the horizon radius with the photon sphere radius. The photon sphere sits at 1.5 times the Schwarzschild radius, about 19 million km for Sagittarius A*. Light can orbit there, but it's not the horizon. Mixing the two inflates the answer by 50 percent, and the mistake shows up in plenty of explainer articles and classroom slides.
The second error is confusing the radius with the shadow radius. The observed shadow is about 2.6 times the Schwarzschild radius in apparent angular size, roughly 33 million km equivalent, because gravitational lensing magnifies the silhouette. If someone quotes the "size of the black hole" from the EHT image and gives a radius near 33 million km, they've described the shadow, not the horizon. Both numbers are real; they answer different questions.
A third failure case involves unit conversion. Plugging 4.3 million into the formula but treating it as kilograms rather than solar masses produces an answer off by 30 orders of magnitude. The shortcut constant only works in solar masses; if you use SI units, you must use 2GM/c² with M in kilograms and G = 6.674×10⁻¹¹ m³ kg⁻¹ s⁻². The result comes out in metres, and you divide by 1,000 for kilometres. Watch the units and the formula is forgiving.
Edge cases matter too. A black hole accreting matter has a slightly larger apparent horizon due to the surrounding plasma, but the mathematical event horizon is unchanged by accretion; it depends only on mass and spin. A binary system's black hole has a horizon distorted by its companion's gravity, but Sagittarius A* is effectively isolated, so the isolated Kerr solution applies. These subtleties don't change the headline number for this object, but they explain why precision claims deserve scepticism.
Rebuilding the Figure From Published Inputs
If you want to verify 12.7 million km yourself, the path is short. Start with the GRAVITY Collaboration's 2020 mass estimate of 4.297 million solar masses, published in Astronomy & Astrophysics. Multiply by 2.953 km per solar mass. The product is 12.69 million km, which rounds to 12.7 million. That's the whole calculation, and anyone with a calculator can reproduce it in under a minute.
To propagate the uncertainty, take the published mass error of roughly 0.1 percent, which is about 4,300 solar masses. Multiply by 2.953 and you get about 12,700 km of radius uncertainty from the mass alone. The distance uncertainty adds a comparable amount. So a defensible error bar is roughly plus or minus 25,000 to 30,000 km on the 12.7 million km figure, small in relative terms but larger than the Earth's diameter.
The spin uncertainty dominates the physical answer, though. Because spin isn't pinned, the true horizon could sit anywhere from about 8.6 to 12.7 million km depending on the actual rotation rate. That factor-of-1.5 range swamps the mass and distance errors. Any article that quotes 12.7 million km as the measured horizon radius without mentioning spin is overstating what's known.
For a back-of-envelope check, use the light-crossing time. A 12.7 million km radius corresponds to 42 light-seconds. If you see a source quote a light-crossing time much longer, they're probably describing the photon ring or the shadow. If they quote much shorter, they've likely made a unit error. The light-crossing time is the fastest cross-check available and requires no formula.
FAQ
Is the event horizon radius of Sagittarius A* exactly 12.7 million km?
No. The 12.7 million km figure is the Schwarzschild radius for a non-rotating black hole of 4.3 million solar masses, and it's an upper bound. A rotating black hole's horizon shrinks by up to half, so the true horizon could be as small as roughly 6.35 million km at maximum spin. Current spin estimates for Sagittarius A* are too uncertain to narrow that range much below 8.6 to 12.7 million km. Quote the number with that caveat attached.
How does the EHT image relate to the horizon radius?
The EHT image shows a ring about 51.8 microarcseconds across, corresponding to a radius near 3.3 billion km at the object's distance. That's roughly 260 times the Schwarzschild radius. The ring is the photon ring, shaped by light orbiting at 1.5 times the Schwarzschild radius and then lensed outward. The event horizon itself emits no light and cannot be imaged, so the picture shows structures well outside the horizon.
What is the horizon radius in astronomical units?
Sagittarius A*'s Schwarzschild radius of 12.7 million km equals about 0.085 astronomical units. For comparison, Mercury orbits at 0.39 AU and the Sun's radius is 0.00465 AU. The horizon would sit inside Mercury's orbit if the black hole replaced the Sun. M87*'s horizon, by contrast, spans about 128 AU, extending far beyond Pluto's orbit at roughly 39 AU.
Does the horizon radius change over time?
For an isolated black hole, no. The horizon radius depends only on mass and spin, and Sagittarius A* accretes very little material, so its mass is essentially constant on human timescales. The accretion rate is far too low to measurably change the radius in a human lifetime. In a merger, the radius would jump as the mass grows, but no merger is expected here.
Why do some sources quote 24 million km?
Some sources confuse the diameter with the radius and double the figure, while others include the photon sphere at 1.5 times the Schwarzschild radius. A 24 million km figure roughly matches 1.9 times the Schwarzschild radius, which could come from mixing the shadow size, the photon sphere, or a diameter calculation. Check whether the source states radius, diameter, or shadow size before comparing numbers.
How does the size compare with a stellar-mass black hole?
A stellar-mass black hole of 10 solar masses has a Schwarzschild radius of about 29.5 km, roughly the length of Manhattan. Sagittarius A* at 4.3 million solar masses is 430,000 times more massive, so its horizon radius is 430,000 times larger: 12.7 million km versus 29.5 km. The linear scaling means you can move between any two black holes by comparing masses alone.
Can a spacecraft cross the horizon, and what would that involve?
In principle, yes, but no human mission is remotely feasible. The nearest horizon is trillions of kilometres away, and crossing it would require surviving extreme tidal forces, which for a supermassive black hole like Sagittarius A* are actually gentler at the horizon than for a stellar-mass hole. The practical barrier is distance and propulsion, not the horizon itself. No spacecraft has come close to any black hole.