What Happens Inside a Black Hole — Beyond the Event Horizon

The event horizon of a black hole is one of the most dramatic boundaries in the universe: a surface from which nothing, not even light, can return. Cross it and you are committed. But what happens after you cross it? The answer involves one of the strangest swaps in physics. Inside a black hole, the roles of time and space are rearranged so that the future is no longer "forward in time" in the everyday sense but "inward, toward the centre." And at the centre, our best theory of gravity stops giving sensible answers.

This article is about the interior specifically. If you want the general picture first (how black holes form from collapsing stars, how they grow and how we detect them), start with black holes explained, from formation to Hawking radiation. Here we go past the horizon, and we keep a careful line between three kinds of statement: what general relativity firmly predicts, what is a reasonable but untested extrapolation, and what is genuinely open research.

Key takeaways

• Crossing the event horizon of a large black hole would feel like nothing in particular: there is no physical barrier there.
• Inside the horizon, the radial direction toward the singularity behaves like time. You cannot avoid moving toward it any more than you can avoid tomorrow.
• Tidal forces (spaghettification) would tear apart anything falling into a stellar-mass black hole before it reaches the horizon; at a supermassive black hole's horizon they are gentle.
• The maximum time anyone can survive inside a non-rotating black hole is about 15 microseconds per solar mass: about a minute for the Milky Way's black hole, roughly a day for M87*.
• The singularity is where general relativity predicts infinite curvature. Most physicists read this as a sign the theory breaks down, not as a literal infinity.
• The information paradox and the firewall paradox are unresolved. Calculations since 2019 suggest information does escape, but how the interior fits into that picture is still debated.

Three levels of confidence

Physics writing about black hole interiors often blurs very different kinds of claim. This table sorts the main ideas in this article by how solid they are.

Claim	Status	Basis
Black holes exist and have event horizons	Established by observation	Gravitational waves (2015 onward), Event Horizon Telescope images (2019, 2022), stellar orbits at the galactic centre
Nothing inside the horizon can reach the outside	Established within general relativity	Follows directly from the geometry; tested indirectly
The singularity is in the infalling observer's future	Established within general relativity	Exact solutions plus Penrose's 1965 singularity theorem
Tidal forces scale with 1/M² at the horizon	Established (classical calculation)	Newtonian and relativistic tidal formulas agree in form
Black holes emit Hawking radiation	Strong theoretical consensus, not observed	Quantum field theory on curved spacetime (Hawking, 1974)
The singularity is literally infinite	Widely doubted	Expected to be replaced by quantum gravity effects
Information escapes in Hawking radiation	Growing theoretical support (2019 onward)	Page curve and "island" calculations, mostly in simplified models
Firewalls exist at the horizon	Open, minority view today	AMPS argument (2012); many proposed escapes
A bounce into a new universe	Speculative	Some quantum gravity models; no observational test yet

Everything above the line of "widely doubted" is mainstream physics. Everything below it is research territory.

The event horizon: what it is and isn't

The event horizon is not a physical surface. There is no wall, no shell of matter, no sudden change in density. In principle you could drift across it without noticing anything locally. It is simply the boundary from which escape becomes geometrically impossible. NASA's own description puts it plainly: inside this boundary, the velocity needed to escape exceeds the speed of light (NASA Science, black hole anatomy).

For a non-rotating (Schwarzschild) black hole, the event horizon is a sphere at the Schwarzschild radius:

r_s = 2GM/c²

Plugging in the numbers gives about 2.95 km per solar mass. A black hole with the Sun's mass would have a horizon about 3 km in radius. The supermassive black hole at the centre of our galaxy, Sagittarius A* (Sgr A*), is about four million times the mass of the Sun, according to the Event Horizon Telescope collaboration that imaged it in May 2022 (ICRR, University of Tokyo, EHT press release). That puts its horizon radius at roughly 12 million km, about 17 times the radius of the Sun. M87*, the black hole in the first EHT image of 2019, is about 6.5 billion solar masses; its horizon radius works out to roughly 19 billion km, larger than Pluto's orbit. The story of how those images were assembled from a planet-sized network of radio dishes is told in the first image of a black hole.

Real black holes almost certainly spin, and a spinning (Kerr) black hole has a more complicated structure: an outer event horizon, a region outside it called the ergosphere where nothing can stay still relative to distant stars, and a second, inner horizon deeper inside. We will come back to that inner horizon, because it changes the interior story considerably.

Spaghettification: tidal forces that kill you first

Before worrying about what happens inside, there is the matter of getting there intact.

Tidal force is the difference in gravitational pull between your head and your feet. Near a mass M, at distance r, the stretching acceleration across a body of length L is roughly 2GML/r³. At the horizon, r = 2GM/c², and the formula simplifies to something that falls off as 1/M². The bigger the black hole, the gentler its horizon.

That has a counter-intuitive consequence. Here is the approximate head-to-foot stretch for a 2-metre-tall person at the horizon of three black holes, compared with Earth's own tidal stretch (all values calculated from the formula above, so treat them as order-of-magnitude):

Black hole	Mass (solar masses)	Horizon radius	Tidal stretch at horizon (2 m body)
Typical stellar-mass	10	~30 km	~2 × 10⁸ m/s² (about 20 million g)
Sgr A* (Milky Way)	~4 million	~12 million km	~10⁻³ m/s² (about a ten-thousandth of g)
M87*	~6.5 billion	~19 billion km	~5 × 10⁻¹⁰ m/s² (imperceptible)
For comparison: Earth's surface	—	—	~6 × 10⁻⁶ m/s²

At a 10-solar-mass black hole, the stretching would pull a human apart into a thin stream of matter well before the horizon is reached. This is spaghettification, and it is the fate of stars torn apart in "tidal disruption events" that astronomers observe as bright flares in distant galaxies. At Sgr A*, the stretch at the horizon is around a thousandth of a metre per second squared: noticeable to sensitive instruments, not to you. At M87* it is far weaker than the stretch Earth's own gravity already applies across your body right now.

So if you want to see the inside of a black hole, choose a big one. The comparison of stellar-mass and supermassive black holes covers why the two populations differ so much in mass, and why there seems to be a gap between them.

Crossing the event horizon: two incompatible-looking stories

General relativity tells two stories about the same fall, and both are correct.

From the perspective of a distant observer watching you fall in:

• As you approach the horizon, light from you is redshifted to longer and longer wavelengths and grows dimmer.
• Your clocks appear to run slower and slower; you seem to approach the horizon without ever quite reaching it.
• You eventually fade from view as the few photons that still escape are stretched beyond detection.

From your own perspective, falling in:

• You cross the horizon in a finite amount of your own time (your "proper time").
• For a large enough black hole, you feel nothing special at the moment of crossing.
• Looking back, you can still see the outside universe; light from outside keeps falling in after you.

These two accounts do not contradict each other, because they use different clocks. The distant observer's time coordinate becomes useless at the horizon, in the same way that longitude becomes meaningless at the North Pole. The "freezing" is a feature of how signals reach the outside, not of what happens to you. Time dilation of this kind is not exotic; much milder versions are measured every day, as explained in how GPS uses Einstein's relativity.

Inside: where the singularity becomes a moment, not a place

Here is the part that breaks intuition. In ordinary spacetime, you move inevitably forward in time, but you can choose to move in any spatial direction. Inside the event horizon of a Schwarzschild black hole, the geometry changes so that decreasing radius becomes the direction of the future. Light cones, which describe every path a signal or traveller could possibly take, tip over so far that all of them point inward.

A cleaner way to say this, and one closer to how relativists put it, is that the singularity is not a point in space in front of you. It is a moment in your future. Asking how to steer away from it is like asking how to steer away from next Tuesday.

Practical consequences, all following directly from the classical solution:

• There is no hovering. Outside the horizon a rocket can, in principle, hold its position. Inside, no amount of thrust can keep you at a constant radius.
• Firing your engines makes it worse. The longest possible survival time comes from falling freely from rest at the horizon. Accelerating hard in any direction, including "outward", shortens your remaining proper time. This is a known and slightly perverse result of the Schwarzschild geometry.
• You cannot see the singularity coming. Because it lies in your future rather than ahead of you in space, light from it never reaches you before you reach it.

How long do you have?

For a non-rotating black hole, the maximum proper time from horizon to singularity is π GM/c³. Since GM/c³ for one solar mass is about 4.9 microseconds, the limit is roughly 15.5 microseconds per solar mass. Scaling up:

Black hole	Mass	Maximum time from horizon to singularity
Stellar-mass	10 solar masses	~0.15 milliseconds
Sgr A*	~4 million solar masses	~1 minute
M87*	~6.5 billion solar masses	~28 hours

These figures are calculated from the textbook formula and the masses cited above. They assume a non-rotating black hole and free fall from rest at the horizon; real black holes spin, which complicates the picture.

What would you see on the way down?

Ignoring firewalls for a moment (more on them below), an infalling observer in a supermassive black hole has time to look around. Ray-tracing studies of the Schwarzschild interior give a fairly consistent picture:

• The outside universe does not shrink to a point overhead, as older popular accounts said. Instead, the region of sky containing the outside universe stays large, and its light becomes distorted, partly blueshifted directly behind you and redshifted toward the edges.
• Light that orbited the black hole on the way in produces multiple, smeared images of the same stars.
• Because everything else falling in shares your future, nearby debris remains visible, though increasingly stretched.
• Close to the singularity, tidal forces grow without limit. Spaghettification happens even inside a supermassive black hole; it is just postponed to the very end.

There is one more consequence worth noting: you never get to see what is "inside" in the way you might imagine. Everything you can observe was either with you when you crossed or fell in after you. The interior holds no stored image of earlier victims.

Rotating black holes: the inner horizon problem

The Schwarzschild solution is mathematically simple but physically idealised. Real black holes inherit spin from the stars and gas that formed them, so they are better described by the Kerr solution, found by Roy Kerr in 1963.

Inside a Kerr black hole, things change in an important way. Between the outer event horizon and the centre lies an inner horizon (also called a Cauchy horizon). Below it, the mathematics of the idealised solution no longer forces you into the singularity; in the exact Kerr geometry, the singularity is a ring rather than a point, and some paths miss it entirely. Taken at face value, the equations even connect to other regions of spacetime, which is where many wormhole and "white hole exit" stories come from. The explainer on wormholes covers why those connections are not considered traversable.

Most relativists do not believe the inner region of the exact Kerr solution is physical. The reason is an instability first analysed by Eric Poisson and Werner Israel in 1989–90 and usually called mass inflation: radiation and matter falling in behind you pile up at the inner horizon, blueshifted to enormous energies, and the effective mass-energy measured there grows without bound. The usual expectation is that a real, perturbed rotating black hole replaces the smooth inner horizon with a violent curvature singularity, though there is ongoing mathematical work on exactly how strong it is and whether an observer would be destroyed immediately. This is an active research area rather than a settled result.

The singularity: where general relativity runs out

General relativity predicts that at the centre of a Schwarzschild black hole there is a singularity, where spacetime curvature, and with it the tidal force, becomes infinite.

For decades this could be dismissed as an artefact of perfect symmetry: maybe a real, lumpy collapsing star would miss its own centre and avoid the infinity. In January 1965, Roger Penrose showed that this escape does not work. His paper "Gravitational Collapse and Space-Time Singularities", published in Physical Review Letters, proved that once a "trapped surface" forms, a singularity follows under very general conditions, regardless of symmetry. That work earned Penrose half of the 2020 Nobel Prize in Physics, cited "for the discovery that black hole formation is a robust prediction of the general theory of relativity". His career, and his more controversial later ideas, are covered in our profile of Roger Penrose.

Precision matters here. Penrose's theorem says spacetime is incomplete: some paths through it end after a finite time. It does not say an infinite density literally exists. Most physicists interpret the result as a signal that general relativity is being pushed beyond its domain of validity. NASA's black hole pages make the same point: the singularity "may be either a physical structure or a purely mathematical one", and astronomers do not know which.

Where should the classical theory fail? The natural scale is the Planck length, about 1.6 × 10⁻³⁵ m, where quantum effects of gravity are expected to become as important as classical curvature. The corresponding Planck density is enormous, on the order of 10⁹⁶ kg/m³. Long before an infalling observer reached that regime they would be destroyed, but the matter that fell in would, according to general relativity, continue toward it.

What actually happens there is unknown because we do not have a complete theory of quantum gravity. Different approaches give different hints:

• String theory suggests that in some cases the singularity is replaced by a structure of strings and branes; in certain versions ("fuzzball" proposals) there is no conventional interior at all.
• Loop quantum gravity models often find a "bounce", where collapse halts at extreme density and turns around, possibly connecting to a white hole or a new expanding region.
• Other approaches propose regular cores, where the interior settles into a finite-density state instead of collapsing to a point.

None of these has been tested observationally, and none is a consensus answer.

Hawking radiation: why the interior matters to the outside

For a long time the interior could be treated as a sealed box: whatever happens inside stays inside. Stephen Hawking's 1974 calculation, published in Nature, broke that comfortable separation. Applying quantum field theory to the curved spacetime outside a horizon, he found that black holes should emit faint thermal radiation and slowly lose mass. Our full explainer on Hawking radiation goes into the mechanism.

The temperature is tiny for any black hole we know of. It is inversely proportional to mass, about 60 billionths of a kelvin for a black hole of one solar mass. That is far colder than the 2.7 K cosmic microwave background, so every known black hole currently absorbs more energy than it emits and is still growing. Evaporation would only begin in earnest in a far-future universe that has cooled further, and the lifetime of a solar-mass black hole is on the order of 10⁶⁷ years (a standard textbook estimate). No Hawking radiation has ever been detected from an astrophysical black hole.

Still, in principle, black holes do not last forever. And that is where the trouble starts.

The information paradox

Quantum mechanics has a strict rule, known as unitarity: information is never fundamentally destroyed. If you knew the exact final state of a system, you could in principle reconstruct its initial state. Burning a book scrambles its information into smoke and light, but does not delete it.

Hawking radiation, as Hawking first calculated it, is exactly thermal, which means it is random and carries no detailed record of what fell in. If a black hole formed from a collapsing star, or from a library of encyclopedias, it would evaporate into the same featureless glow. When the black hole is gone, the information appears to be gone too. That violates unitarity.

This is the black hole information paradox, and it has driven a large part of theoretical physics since the mid-1970s. Hawking himself initially argued that information really is lost; he later changed his position and publicly conceded a bet on the question in 2004. Our profile of Stephen Hawking recounts the wagers.

Proposed resolutions have included:

• Information comes out in subtle correlations in the Hawking radiation, so the radiation only looks thermal if you examine it piece by piece.
• Information comes out at the very end, when the black hole is Planck-sized and quantum gravity dominates.
• Remnants: evaporation stops, leaving a tiny object that stores everything.
• Complementarity: an outside observer sees information stored and re-emitted near the horizon, while an infalling observer sees it pass smoothly inside, and no single observer can ever compare the two accounts.

The Page curve and the "islands" breakthrough

In 1993, Don Page pointed out what a unitary answer would have to look like. Track the entanglement entropy of the radiation over time. At first it rises, as each emitted quantum is entangled with partners left behind. If information is preserved, the curve must turn over at roughly the halfway point of evaporation (now called the Page time) and fall back to zero by the end. Hawking's original calculation gives a curve that only rises.

In May 2019, two groups posted papers within a day of each other showing how to get the turnover. Geoffrey Penington argued that evaporating black holes in a holographic setting undergo a phase transition at the Page time, which yields the Page curve and lets interior information be recovered from the radiation (arXiv:1905.08255). Ahmed Almheiri, Netta Engelhardt, Donald Marolf and Henry Maxfield reached a matching result in a simplified two-dimensional gravity model (arXiv:1905.08762). Follow-up work introduced the language of "islands": after the Page time, a region inside the black hole is, in a precise quantum sense, encoded in the radiation outside.

This is the most important development on the topic in decades, but it comes with caveats. Most of the calculations are done in simplified or lower-dimensional models, often relying on the holographic duality between gravity and quantum field theory. They show that the radiation can carry the information, without a detailed microscopic account of how individual quanta encode it. And they say surprisingly little that is intuitive about what the interior "is" for someone who has fallen in.

The firewall paradox

In July 2012, Ahmed Almheiri, Donald Marolf, Joseph Polchinski and James Sully (AMPS) sharpened the problem into an argument that unsettled the field. In their paper "Black Holes: Complementarity or Firewalls?" they argued that three reasonable claims cannot all be true together (arXiv:1207.3123):

1. Hawking radiation is pure: information escapes, so quantum mechanics is respected.
2. Ordinary physics works near the horizon: low-energy effective field theory is valid just outside it.
3. No drama: an infalling observer notices nothing unusual when crossing the horizon.

The core of the argument rests on a rule called monogamy of entanglement. A quantum system cannot be fully entangled with two separate partners at once (the basics are in quantum entanglement explained). For information to come out, late Hawking quanta must be entangled with earlier radiation. For the horizon to be smooth, the same quanta must be entangled with partners just inside. Both cannot hold.

The authors suggested that the most conservative fix was to give up the third claim: the infalling observer burns up at the horizon, hitting a "firewall" of high-energy quanta. If that is right, the gentle crossing described earlier in this article would be wrong for any black hole old enough to have passed its Page time.

Most theorists find a firewall hard to accept, because it would mean general relativity fails badly in a region where curvature is low. Proposed escapes include the "ER = EPR" conjecture of Juan Maldacena and Leonard Susskind (entanglement and wormhole-like connections as two sides of the same thing), state-dependent reconstructions of the interior, and the island picture itself, which reorganises which degrees of freedom are really independent. There is no agreed resolution as of 2026, though the consensus has drifted toward believing horizons are smooth and that the paradox is resolved by subtler quantum structure rather than literal fire.

Common misconceptions

"The singularity is a point at the centre, like a pea in a peach." In the non-rotating solution, the singularity is better thought of as a moment in time for anyone inside, not a location you could point at.

"Time stops at the event horizon." Time stops only in the bookkeeping of a distant observer. The infalling observer's own clock ticks normally.

"Black holes suck things in like a vacuum cleaner." Far from a black hole, its gravity is the same as that of any object of the same mass. If the Sun were replaced by a one-solar-mass black hole, Earth's orbit would not change.

"We have seen inside a black hole." The Event Horizon Telescope images show the shadow and the glowing gas just outside the horizon. Nothing from inside can reach us, by definition.

"Black holes are holes in space." In general relativity they are regions of extremely curved spacetime, not punctures in it, and they have mass, spin and (in principle) charge that can be measured from outside.

Frequently asked questions

Q: Could there be a universe inside a black hole? A: Some speculative models propose that collapse inside a black hole ends in a bounce, from which a new expanding region of spacetime could form. This idea appears in some quantum gravity approaches but has no observational support, and there is currently no known way to test it from outside.

Q: Is there anything at the centre of a black hole? A: Nobody knows. General relativity predicts a singularity, but that prediction most likely marks where the theory breaks down rather than a real infinity. Describing the centre needs a working theory of quantum gravity, which does not yet exist.

Q: How long would it take to hit the singularity after crossing the horizon? A: For a non-rotating black hole, at most about 15.5 microseconds per solar mass of the black hole, measured on your own clock. That works out to roughly a minute for Sagittarius A* and about 28 hours for M87*. Firing rockets in any direction only shortens the time.

Q: Can information actually escape via Hawking radiation? A: Calculations published since 2019, including the "island" and Page curve results, strongly suggest that it can, at least in simplified models. Exactly how individual particles of radiation carry the information is still not understood.

Q: What is the difference between a stellar-mass and a supermassive black hole interior? A: The geometry is qualitatively the same, but the scale is not. A stellar-mass black hole destroys an infalling body before the horizon through tidal forces. A supermassive one lets you cross intact, and the time left before the singularity grows in proportion to the mass.

Q: Have we ever seen inside a black hole? A: No. The event horizon is, by definition, the boundary beyond which no signal can escape. Telescopes observe the surrounding gas, the shadow and the gravitational waves from mergers, all of which come from outside the horizon.

Q: Would a firewall really kill you at the horizon? A: If the 2012 firewall argument were correct, yes, for an old enough black hole. Most physicists suspect that subtler quantum effects resolve the paradox and that the horizon is smooth, but the question is not settled.

The bottom line

Crossing a supermassive black hole's event horizon would feel unremarkable: no barrier, no warning sign. But inside, the most profound change in the geometry of spacetime takes place. The singularity becomes part of your future, and no rocket can change that. This much is solid general relativity. What happens at the end of that future, and whether the interior is truly sealed off from the rest of the universe, is not settled. Penrose showed the singularity is unavoidable in classical theory; Hawking showed black holes evaporate; and the information and firewall paradoxes show that our two best theories of nature disagree about what the inside of a black hole even is. The partial answers found since 2019 point toward information escaping, but the interior remains the place where physics, as we currently write it down, stops working.

Sources

• NASA Science, "Black Hole Anatomy": https://science.nasa.gov/universe/black-holes/anatomy/
• Event Horizon Telescope announcement of the Sgr A* image, 12 May 2022 (ICRR, University of Tokyo): https://www.icrr.u-tokyo.ac.jp/en/news/11810/
• Almheiri, Marolf, Polchinski, Sully, "Black Holes: Complementarity or Firewalls?" (2012): https://arxiv.org/abs/1207.3123
• Penington, "Entanglement Wedge Reconstruction and the Information Paradox" (2019): https://arxiv.org/abs/1905.08255
• Almheiri, Engelhardt, Marolf, Maxfield, "The entropy of bulk quantum fields and the entanglement wedge of an evaporating black hole" (2019): https://arxiv.org/abs/1905.08762

Related Reading

• What Are Black Holes Made Of? Peering into the Universe's Darkest Mystery
• What Would Happen If You Fell Into a Black Hole?
• How Black Holes Form — and What Happens Inside Them
• What Are Quasars — The Brightest Objects in the Universe?