How Are Black Holes Formed

Short answer

• Black holes are formed in more than one way, and the route depends on the size of the black hole that results. The best established route is the death of a massive star. When such a star runs out of fuel, its core can no longer hold itself up against gravity, so it collapses. If the collapsed core is massive enough, no known force stops the collapse, and a stellar mass black hole is born. Bigger black holes are built, not simply born in one step. Smaller black holes can merge, gas can feed a growing black hole over time, and dense star clusters can push black holes together. For the largest black holes, astronomers also test two seed ideas: small seeds that grow fast, and heavy seeds that start large when a huge gas cloud collapses directly. A separate idea, primordial black holes, says that some black holes could have formed from dense patches in the very early universe, without any star at all. That last idea remains a hypothesis, not a confirmed observation.

One Question, Several Answers

Asking how black holes are formed sounds like asking for a single recipe. In practice, astronomers give a set of routes, because black holes span an enormous mass range, from a few times the mass of the Sun to millions or billions of times that mass, with a rarer intermediate class in between. No single stellar collapse can plausibly explain the whole range, so the story must explain both birth and growth.

This article is about formation only. It does not repeat a full definition of what a black hole is. For that basic picture, see the companion explainer on what a black hole is.

The safest way to hold the picture in mind is to separate birth from growth. Birth is the moment a black hole first exists. Growth is everything after that. Many arguments about supermassive black holes are really arguments about whether growth alone can do the job in the time available, or whether the starting seed also has to be large.

Stellar Mass Black Holes: The Death of a Massive Star

The most firmly established formation route starts with a star at least about 10 times as massive as the Sun. NASA describes this sequence in its educational material on black holes. Such a star lives fast. It fuses hydrogen into helium, then fuses heavier elements in turn, while the outward push of fusion balances the inward pull of gravity. The balance cannot last forever.

When the core has become essentially iron, the sequence reaches a problem. Fusing iron does not release energy in the way that fusing lighter elements does, so the core loses its support. NASA states that fusion then ceases, and in less than a second the star begins the final phase of gravitational collapse. The outer layers are driven outward in a supernova explosion, while the core itself is crushed inward.

What remains depends on how massive the crushed core is. If the core is not massive enough, it can settle as a neutron star, an extremely dense ball made mostly of neutrons. If the original star was very massive, about 10 or more times the mass of the Sun, NASA explains that even neutrons cannot withstand the collapse, and a black hole forms. The resulting stellar mass black holes typically have masses from a few to a few tens of solar masses, according to the same NASA material.

This route explains why stellar mass black holes are the most studied kind. Massive stars are rare, but the universe has made a great many over billions of years, and each suitable death can leave one black hole behind, alone or in a binary system.

Note that the supernova itself is not the black hole. The supernova is the explosion of the outer layers, while the black hole is what the core becomes when its collapse cannot be stopped.

Why Not Every Star Qualifies

Most stars will never become black holes, and the reason is mass. A star like the Sun does not have enough mass for its core to collapse past the neutron star stage, let alone to a black hole. NASA states the present day minimum plainly in its material on primordial black holes: the smallest black holes that form nowadays are born when a massive star runs out of fuel, and at least eight times the mass of the Sun is required. Lighter stars end as white dwarfs or neutron stars instead. The 10-times figure comes from NASA’s stellar life-cycle account and the eight-times figure from NASA’s primordial-black-hole account; both are approximate source thresholds for the same mass rule, not two different rules.

That threshold is a starting mass for the star, not the final mass of the black hole. A massive star loses a large amount of material during its life, through stellar winds and through the supernova itself. The black hole that remains therefore weighs less than the star that started the process. Astronomers must track both numbers: how massive the star was at birth, how much mass it kept until collapse, and how massive the remaining core was at the moment of collapse. A star that begins above the threshold can still fail to leave a black hole if it sheds too much mass before the end.

Composition also matters. Stars with more heavy elements tend to drive stronger winds and lose more mass, while early universe stars, formed from almost pure hydrogen and helium, may have kept more mass and left heavier remnants.

Rotation and binary companions add variation, which is why astronomers give ranges rather than a single exact cutoff.

When There Is No Supernova: Failed Supernovae

Not every massive star ends with a bright explosion. Some appear to collapse quietly, in what astronomers call a failed supernova. NASA reported one such case in 2017, when a team from Ohio State University watched a massive star disappear. The star was 25 times as massive as the Sun, and it should have exploded in a very bright supernova. Instead, it fizzled out, and the researchers concluded, by a process of elimination, that it had become a black hole.

The researchers were careful about the size of the claim. Christopher Kochanek of Ohio State University said such failures could explain why astronomers rarely see supernovae from the most massive stars. Scott Adams, a former Ohio State student who had recently earned his doctorate doing this work, gave a preliminary estimate that 10 to 30 percent of massive stars might die as failed supernovae, based on one likely case in seven years, during which six normal supernovae occurred in the monitored galaxies. That estimate is preliminary, not settled.

A second, more recent case strengthens the picture. NASA reported in February 2026 that archival data from the NEOWISE mission, together with other space and ground based observations from 2005 to 2023, had pieced together the fate of a dying star in the Andromeda galaxy named M31-2014-DS1, about 2.5 million light years from Earth. The star brightened in infrared light in 2014, then dimmed in visible light by more than a factor of 10,000 by 2023. The researchers concluded that the star shed its outermost layers as its inner core ran out of fuel, and that most of the stellar material then collapsed under gravity into a black hole, because a weak shockwave could not push the material outward.

Failed supernovae matter because a violent explosion blows much mass away, while a quiet collapse lets more mass stay in place, making a heavier black hole easier to form.

Mergers: How Black Holes Build Bigger Black Holes

Once black holes exist, they can combine. Two black holes in a binary system slowly spiral toward each other as they radiate gravitational waves, ripples in spacetime predicted by general relativity. When they finally merge, they form a single, heavier black hole. Some of the combined mass is carried away as gravitational wave energy, so the remnant weighs slightly less than the sum of the two original black holes, but it is still much heavier than either one alone.

Gravitational wave observatories have turned this growth route from theory into observation. The LIGO Scientific Collaboration and its partners have recorded close to 300 black hole mergers by the time of their GW231123 summary. Each detection gives the masses of the merging black holes and the mass of the remnant, so astronomers can test whether a given remnant could plausibly have come from ordinary stellar collapse, or whether it must have had a more complex history.

Mergers also let black holes grow where gas is scarce, provided the environment is crowded enough for black holes to meet, and provided the merged remnant is not kicked out by the recoil of the merger.

Keep birth and growth separate here: a merger creates a larger black hole from two that already existed, so a complete story still needs a birth route for the first black holes in the chain.

Intermediate Mass Black Holes: The Missing Middle

Intermediate mass black holes sit between the stellar and supermassive classes, heavier than those formed by a single stellar collapse, but much lighter than the black holes at galaxy centers. They have been hard to confirm, because they are not massive enough to dominate a whole galaxy, and not common enough to turn up routinely in stellar surveys. Their formation is therefore one of the most interesting open problems in this field, since it tests whether the mass ladder from stellar to supermassive is continuous.

The LIGO Scientific Collaboration summary for GW231123 gives a clear example of how a merger can produce an intermediate mass black hole. The signal, detected on November 23, 2023, came from two black holes of approximately 137 and 103 times the mass of the Sun. Their total mass was likely between 190 and 265 solar masses, and the merger produced a black hole likely between 182 and 251 solar masses. That remnant falls in the intermediate mass category, and the collaboration describes the remnants of GW231123 and the earlier GW190521 as the clearest gravitational wave detections of these elusive middle sized black holes.

GW231123 is also interesting because its component black holes challenge simple stellar evolution. Current theories suggest that black holes with masses between approximately 60 and 130 solar masses should be rare or nonexistent, a range called the black hole mass gap, because certain kinds of explosions either tear heavy stars apart or eject much of their mass before collapse. The LIGO summary notes that the lighter black hole in GW231123 has an 83 percent chance of falling inside that gap, while the heavier one has a 26 percent chance. These percentages depend on the waveform models used to interpret the very short signal; LIGO notes that the signal pushes those models to their limits. One possibility raised by the collaboration is that one or both of these black holes formed through earlier mergers of smaller black holes, in a dense environment such as a nuclear star cluster or an active galactic nucleus, where black holes are more likely to collide.

A second proposed route is runaway collisions in a dense star cluster, where massive stars collide and merge repeatedly, building a very massive star that then collapses into a heavier black hole. This route is a hypothesis supported by models and indirect evidence, not a process watched from start to finish, and the full history of any individual intermediate mass black hole is usually uncertain.

Supermassive Black Holes: The Seed Problem

Supermassive black holes pose a timing problem. Astronomers find black holes millions or billions of times the mass of the Sun less than a billion years after the Big Bang. If they started as small stellar mass black holes, they would have to grow extremely fast to reach such masses so early, which strains the simplest growth models, so astronomers test whether at least some started large.

The two seed ideas are usually called light seeds and heavy seeds. In the light seed picture, a black hole seed of about 10 to 100 times the mass of the Sun forms from the collapse of a massive star, then grows by mergers and by pulling in gas. In the heavy seed picture, a much larger seed forms directly from the collapse of a giant gas cloud, skipping the star stage entirely. The European Space Agency, reporting a 2016 study by an Italian team, described two seed candidates seen less than a billion years after the Big Bang with an initial mass of about 100,000 times the Sun. The researchers stressed that their discovery would explain the birth of monster black holes only if confirmed, and that further observations were needed to confirm the true nature of the candidates.

NASA reported the same study, with lead author Fabio Pacucci saying the team had found evidence that seeds can form directly from the collapse of a giant gas cloud, skipping intermediate steps. The key word is evidence, not proof: matching a prediction is not the same as confirming the object.

A newer observation from the James Webb Space Telescope sharpens the question. NASA reported in May 2026 that researchers had studied Abell2744-QSO1, known as QSO1, a small object that existed just 700 million years after the Big Bang. Using Webb, the team measured the motion of hydrogen gas around its black hole and found Keplerian motion, gas orbiting a central point in the same way that planets orbit the Sun. That measurement let them calculate the black hole mass directly, at roughly 50 million solar masses, and they found that the black hole makes up at minimum two thirds of the total mass of QSO1. The gas is also almost entirely hydrogen and helium, with a metallicity less than 0.5 percent of the Sun, which means very little material from earlier stars is present.

The team inferred from the outsized mass that QSO1 cannot have formed gradually from much smaller black holes merging and feeding, and that it is evidence for primordial or direct collapse black holes, theorized but not confirmed. QSO1 is therefore strong evidence for a heavy seed, not a settled identification of one channel, since both a primordial origin and a later direct collapse origin remain possibilities.

Direct collapse also needs an explanation for why the gas cloud does not simply fragment into stars. The heavy seed theory proposes a much larger black hole forming directly from a large gas cloud, and its central difficulty is that gas clouds like to form stars as they collapse rather than a black hole. Proposed mechanisms include radiation keeping the gas hot so it cannot cool and fragment, and rapid gas inflows with turbulence during a galaxy merger that keep the inner gas from fragmenting. None is confirmed as the general answer, so direct collapse remains a theory with supporting candidates, not an established route on the same footing as stellar collapse.

Primordial Black Holes: A Hypothesis From the Early Universe

Primordial black holes are different from every other kind discussed so far, because they would not come from stars, clusters, or gas clouds in galaxies. They would come from the universe itself, in its first chaotic moments. NASA explains the idea in its material on primordial black holes: scientists suspect that smaller primordial black holes, including some with masses similar to Earth, could have formed in the first moments of the early universe, when regions denser than their surroundings collapsed directly.

The mechanism depends on the extreme conditions of the infant universe. During inflation, space expanded faster than the speed of light, and a patch dense enough could have collapsed under gravity without waiting for any star to form. With no massive star required, there is no eight solar mass minimum, so a primordial black hole could in principle be as light as Earth, or far lighter.

NASA also states the limits of the idea. Theory predicts that the smallest primordial black holes should evaporate before the universe reaches its current age, while those with masses similar to Earth could have survived. An Earth mass primordial black hole would have an event horizon about as wide as a United States dime coin, which shows how hard such an object would be to find: it would be tiny, dark, and detectable only through its gravity, for example by gravitational lensing or by its effect on nearby matter.

No primordial black hole has been confirmed. The idea remains valuable because it makes testable predictions, but a hypothesis that could explain an observation is not a confirmed explanation. Until one is identified by a distinctive signature that rules out other origins, it stays in the hypothesis column, separate from stellar collapse, which is observed.

How Astronomers Test a Formation Story

Formation theories are tested by the fingerprints each route should leave. Stellar collapse should leave supernova remnants, neutron stars in some cases, and black holes whose masses match stellar models. A star disappearing without a bright explosion supports quiet collapse, while gravitational wave masses in the mass gap challenge simple stellar models and point toward earlier mergers.

Supermassive formation is tested mainly by looking back in time. Webb observes galaxies as they were less than a billion years after the Big Bang and measures how massive their black holes already were. If early black holes are already enormous in small, pristine host galaxies, as with QSO1, slow growth from small seeds becomes harder to defend. If early black holes grow in step with their hosts, light seeds with rapid feeding remain plausible. Both ideas are still live.

Dense clusters are tested by measuring star motions near cluster centers and by searching for gravitational wave signals from cluster mergers. These tests are indirect, since a concentrated group of smaller remnants can mimic a single larger black hole, so each candidate needs careful follow up.

Surveys for failed supernovae test the stellar route directly. If failed supernovae are common, a matching fraction of massive stars should be observed to vanish without exploding. The preliminary estimate of 10 to 30 percent is a starting point for that test, not a final answer.

Open Questions

Several questions remain genuinely open, and a good formation article should name them rather than smoothing them over.

First, what fraction of massive stars die as failed supernovae, and does that fraction depend on mass, composition, or environment? The two cases described here show that quiet collapse happens, but they do not yet fix how common it is.

Second, how do intermediate mass black holes form most often: by repeated mergers in clusters, by the collapse of a runaway merger star, by a direct collapse process, or by a mix of routes? GW231123 shows that mergers can produce one, but it does not show that mergers are the only route.

Third, did supermassive black holes start as light seeds, heavy seeds, or both? The evidence from QSO1 and from the ESA seed candidates favors at least some heavy seeds, but it does not rule out light seeds in other galaxies, and it does not yet identify a confirmed direct collapse object beyond reasonable doubt.

Fourth, do primordial black holes exist at all, and if so, in what mass range and in what numbers? A confirmed primordial black hole would rewrite the early chapters of this story, because it would show that black holes can form without stars. Until then, the idea stays as a hypothesis with clear predictions and no confirmed example.

Fifth, what prevents a direct collapse gas cloud from fragmenting into stars? Until astronomers can show a mechanism that works under realistic conditions, and observe it in action, direct collapse will remain a promising theory rather than a settled process.

These questions do not mean astronomers know nothing about formation. The stellar route is well established, mergers are observed, and the seed problem is now constrained by real early universe data. New observations from Webb, gravitational wave detectors, and failed supernova surveys are likely to move the boundary between established and tested next.

Frequently Asked Questions

How are black holes formed in simple terms?

A black hole forms when enough mass is squeezed into a small enough space that gravity wins over every other force. For stellar mass black holes, that happens when a massive star runs out of fuel and its core collapses. For larger black holes, smaller black holes merge, gas feeds a growing black hole, or, in a theory still being tested, a huge gas cloud collapses directly.

Can any star become a black hole?

No. NASA states that at least about eight times the mass of the Sun is required for the smallest black holes that form in the present day universe. Lighter stars become white dwarfs or neutron stars, and even among massive stars, the outcome depends on how much mass survives until collapse.

Do black holes form from supernovae?

A supernova and a black hole are related but not identical. In the standard route, the outer layers explode as a supernova while the core collapses into a black hole. In a failed supernova, the star collapses without a bright explosion. In both cases, the black hole comes from the collapsing core, not from the explosion.

How are supermassive black holes formed?

There is no single settled answer yet. The leading ideas are light seeds, small black holes that grow by feeding and merging, and heavy seeds, large black holes that form directly from collapsing gas clouds. Evidence from Webb, including QSO1, favors at least some heavy seeds, but no channel is confirmed to the exclusion of the other.

What is a primordial black hole?

A hypothetical black hole that would have formed in the first moments of the universe, from a dense patch collapsing directly, without any star. None has been confirmed. The idea is testable, because such black holes could have masses far below the stellar minimum, but it remains a hypothesis.

Are new black holes still forming today?

Yes. Massive stars are still dying, black holes are still merging, as gravitational wave detections show, and supermassive black holes are still growing by feeding and by galaxy collisions.

Why do astronomers disagree about supermassive black hole formation?

Because the early evidence is still incomplete. Early supermassive black holes appear very massive very quickly, which is hard to explain by slow growth from small seeds, but heavy seed candidates are hard to confirm, and direct collapse needs a mechanism to stop gas fragmenting into stars.

Sources & references

• NASA, Imagine the Universe, The Anatomy of Black Holes, Page 4. Stellar mass black holes from the death of a very massive single star, masses from a few to a few tens of solar masses; stars 10 or more times the mass of the Sun, iron core, collapse in less than a second, neutron star or black hole outcome.
• NASA Science, Collapsing Star Gives Birth to a Black Hole, published May 25, 2017. Star 25 times as massive as the Sun, failed supernova, preliminary estimate of 10 to 30 percent of massive stars dying as failed supernovae.
• NASA Science, Archival Data From NASA NEOWISE Tracks Star Turning Into Black Hole, published February 12, 2026. M31-2014-DS1 in the Andromeda galaxy, 2.5 million light years away, infrared brightening in 2014, dimming by more than a factor of 10,000 by 2023, collapse into a black hole.
• European Space Agency, Science and Technology, Hubble finds clues to the birth of supermassive black holes, May 24, 2016. Two seed candidates less than a billion years after the Big Bang, initial mass about 100,000 times the Sun, light seed and direct collapse theories, confirmation still needed.
• NASA Jet Propulsion Laboratory, NASA Telescopes Find Clues For How Giant Black Holes Formed So Quickly. Same seed study reported by NASA: seeds forming directly from the collapse of a giant gas cloud, skipping intermediate steps, with confirmation still needed.
• NASA Science, NASA Webb Reveals Black Hole That Formed Before Its Galaxy, published May 27, 2026. Abell2744-QSO1, 700 million years after the Big Bang, black hole roughly 50 million solar masses, at minimum two thirds of total mass, metallicity less than 0.5 percent of the Sun, evidence for primordial or direct collapse black holes, theorized but not confirmed.
• NASA Scientific Visualization Studio, Primordial Black Holes, released May 7, 2024. Present day minimum about eight times the mass of the Sun, primordial black holes as a suspected class from the first moments of the universe, Earth mass event horizon about as wide as a United States dime coin, smallest predicted to evaporate, Earth mass ones could survive.
• LIGO Scientific Collaboration, GW231123: The most massive black hole binary detected through gravitational waves. Signal detected November 23, 2023, components approximately 137 and 103 times the mass of the Sun, total 190 to 265 solar masses, remnant 182 to 251 solar masses, intermediate mass category, mass gap 60 to 130 solar masses, earlier mergers as a possible origin.