Picture a highway where wrecked cars never stop moving. Every crash leaves the pieces flying along at full speed, and each piece can cause the next crash. That is the basic worry behind the Kessler syndrome, an idea two NASA scientists put forward in 1978: once enough junk circles Earth, collisions between old satellites and rocket parts could keep making new debris faster than nature cleans it up. It is not wiping out satellites today. But the crowding is real, it is getting worse, and the agencies that track it now say the debris would keep growing even if every launch stopped tomorrow.
Why a tiny piece of junk is so dangerous
On Earth, a loose bolt on the road is a nuisance. In orbit, it is a bullet. Anything circling the planet low enough to be useful, a region called low Earth orbit that runs up to about 2,000 kilometers (1,200 miles) above the ground, is moving at roughly 7 to 8 kilometers per second. That is around 17,000 miles per hour. When two objects on different paths meet, they can hit each other even faster.
At those speeds, size stops being a good guide to danger. NASA’s debris experts say a piece about 10 centimeters (4 inches) across can be enough to shatter a whole satellite. Something the size of a marble can still knock one out of service. Even a fleck of paint can leave a pit in a window.
And when something big does break apart, it does not leave one piece of junk. It leaves hundreds or thousands, each now following its own path around Earth. That is where the chain reaction comes in.
Where the name came from
In the 1970s, Donald Kessler was a NASA scientist in Houston who had studied how rocks in the asteroid belt grind each other down over time. When the official catalog of objects in orbit, kept by NORAD, the North American air defense command, became public in the 1970s, he and his colleague Burton Cour-Palais applied the same math to satellites. Their paper, published in the Journal of Geophysical Research in June 1978, argued that collisions would create fragments, the fragments would raise the odds of more collisions, and a belt of debris could slowly form around Earth. In asteroids, that process takes billions of years. Around Earth, they estimated, it could take decades.
Kessler did not name it after himself. John Gabbard, a NORAD analyst who studied breakups in orbit, started calling it the Kessler syndrome in an interview soon after the paper came out. The phrase caught on after a 1982 Popular Science article.
In 1991, Kessler sharpened the idea into what is basically a tipping point. If the amount of junk at a given height is low, the air at the edge of space drags pieces down faster than collisions make new ones, and the problem fades. Above a certain level, called the critical density, each collision makes more junk than drag can remove. Past that point, the debris keeps multiplying on its own, even with no new launches.
One thing the original paper got right that often gets lost: it did not predict a sudden disaster. It described a slow buildup that could become a significant problem over the following century, and it said sensible rules about launches and old satellites could delay it.
The crashes that already happened
For years, the cascade was mostly a math problem. Then real breakups started filling the catalog.
On January 11, 2007, China tested an anti-satellite missile on one of its own old weather satellites, Fengyun-1C, about 865 kilometers (537 miles) up. It remains one of the worst single debris events on record. NASA counted more than 2,000 trackable fragments and estimated at least 150,000 pieces larger than 1 centimeter. At that height, much of it was expected to stay up for decades, some of it for more than a century.
Two years later came the first accident of its kind. On February 10, 2009, a working American communications satellite, Iridium 33, ran into a dead Russian satellite, Cosmos 2251, about 790 kilometers (490 miles) above northern Siberia. They met at about 11.6 kilometers per second, or roughly 26,000 miles per hour. No one aimed them at each other. Their paths just crossed at the same moment. NASA later counted more than 1,800 fragments about 10 centimeters or larger, and expected some of them to stay in orbit until the end of the century.
On November 15, 2021, Russia destroyed its own dead satellite Kosmos 1408 with a missile, creating about 1,500 tracked pieces. The seven people aboard the International Space Station spent about two hours sheltering inside the spacecraft that had brought them there, in case they had to leave in a hurry. The station was not hit.
Not every breakup involves a collision. Old rocket stages can burst when leftover fuel or a battery fails. In August 2024, the upper stage of a Chinese Long March 6A rocket broke apart, leaving at least 700 tracked fragments. The European Space Agency (ESA) counted more than 3,000 new tracked objects added by breakups that year alone.
How crowded is it up there?
As of 2026, the radars and telescopes that watch the sky regularly track about 47,000 objects in orbit, according to ESA’s Space Debris Office. Close to 16,000 of them are working satellites, and most of the rest are dead satellites, spent rocket bodies, and fragments.
The tracked objects are only the big ones, roughly 10 centimeters and up in low orbit. ESA’s computer models estimate about 1.5 million pieces between 1 and 10 centimeters, and around 230 million between 1 millimeter and 1 centimeter. Nobody can see those directly, but they are out there.
That gap matters because satellites can only dodge what they can see. NASA’s Orbital Debris Program Office has said that steering around tracked objects addresses less than 1 percent of the risk of a mission-ending hit. Most of the danger comes from the small stuff, pieces as small as a millimeter or so in low orbit. Spacecraft that carry people get layered armor for exactly this reason, which we covered in how a thin shield stops space rocks going faster than bullets.
Traffic is also rising fast. ESA’s 2026 report, covering data through the end of 2025, says there were more than 300 launches in 2025 and over 4,000 new payloads put into orbit. On average, about ten objects are launched every day, while a little over three intact satellites or rocket bodies fall back into the atmosphere.
Height decides how long junk sticks around
There is one natural cleaner. Even hundreds of kilometers up, there is a thin trace of air. It slowly drags on everything, and over time, objects sink lower and burn up. How fast that happens depends a lot on height.
Below about 600 kilometers (370 miles), drag works quickly enough that dead satellites and fragments usually come down within a few years. Around 800 to 900 kilometers, where the 2007 missile test and the 2009 collision happened, the air is so thin that debris can stay up for centuries. That is why researchers worry most about that band.
Drag also changes with the Sun. When the Sun is active, it heats and puffs up the upper atmosphere, so junk comes down faster, but it also makes satellites’ positions harder to predict.
Operators are leaning on this. In early 2026, SpaceX said it was lowering about 4,400 of its Starlink internet satellites from roughly 550 kilometers to about 480 kilometers. Fewer other objects fly there, and a satellite that fails at that height falls out of orbit in months instead of years.
What experts are worried about now
Here is the part that has changed. For a long time, the debate was whether the tipping point Kessler described would ever arrive. Now the agencies that track debris talk about it in the present tense, at least for some heights.
ESA’s 2026 Space Environment Report says that even if no new satellites were launched, the amount of debris would keep growing, because breakups are adding fragments faster than they fall out of orbit. Its long-range projections show the runaway effect showing up sooner than in earlier reports. ESA also keeps a single score for how sustainable our use of space is, compared with a safe target set in 2014. In one year, that score jumped from about 4 times the safe level to about 50 times. ESA’s conclusion is that preventing new debris is no longer enough, and that some old junk will have to be removed.
Kessler himself wrote in 2009 that models already suggested parts of low orbit had become unstable in this way. Some recent studies point to the band between about 520 and 1,000 kilometers. Even there, researchers stress that the process is slow, playing out over decades rather than days.
A newer way to measure the strain comes from astronomers Sarah Thiele, Aaron Boley, Samantha Lawler, and colleagues. They call it the CRASH Clock. It asks one question: if every satellite suddenly lost the ability to steer, say after a huge solar storm, how long until the first likely crash? In 2018, before the big internet constellations, their answer was 164 days. For mid-2025, it was 5.5 days. By their latest count, in May 2026, it was about 2.5 days. The authors are clear that this is not a countdown to the Kessler syndrome. It is a measure of how little room for error is left. Starlink satellites already swerve to avoid something about once every two minutes on average, around 300,000 maneuvers in 2025.
What is not happening
It is easy to read all that and picture the movie Gravity, where one explosion sets off a storm of debris that shreds everything in orbit within minutes. That is not how it works, and it is not what is going on.
Satellites are not being knocked out one after another. Nearly 16,000 of them are working right now, and new ones go up almost every day. Accidental collisions between big, intact objects are still rare. The 2009 crash remains the most famous example.
A cascade, if it comes, would not be one dramatic event. It would show up as a slowly rising number of collisions per year at certain heights, plus more swerving, more shielding, more lost satellites, and higher costs. Even a serious crash at a busy height like 550 kilometers would not close it off. Operators would face a few years of extra danger until drag cleared the fragments.
It also would not trap us on the ground. The worst debris is in low orbit. GPS satellites fly about 20,000 kilometers up, and TV and weather satellites in geostationary orbit sit near 36,000 kilometers. A cascade in low orbit would pose little risk to those, or to rockets passing through on the way to the Moon or beyond. Launches would face more risk crossing the crowded band, not a wall.
What would slow it down
Most of the fixes are simple to describe and hard to enforce. The first is to stop leaving dead things in orbit. In low orbit, the target used to be removing a satellite within 25 years of the end of its mission. In 2022, the U.S. Federal Communications Commission cut that to five years for the satellites it licenses, and ESA now uses five years too. ESA says more satellites are being brought down on purpose at the end of their missions each year.
The second is to keep old hardware from blowing up. That means venting leftover fuel and draining batteries once a rocket stage or satellite is done. It also means no more missile tests that smash satellites. In 2022, the United States pledged not to carry out destructive anti-satellite missile tests, and other countries have made the same promise.
The third is to haul away the worst offenders. In 2025, a team led by Darren McKnight of the tracking company LeoLabs updated a list of the 50 most dangerous derelict objects in low orbit, mostly old Soviet and Russian rocket stages. Removing all 50 would cut low orbit’s potential for creating new debris roughly in half, and removing just the top 10 would cut it by about 30 percent. ESA is developing a mission called ClearSpace-1 to test grabbing a dead satellite and pulling it down. No mission has yet removed a large piece of junk from orbit this way.
The bottom line
The Kessler syndrome is a real, decades-old idea, and the physics behind it holds up: in orbit, broken things keep moving, and they break other things. It is not happening as a sudden disaster, and satellites are not dropping out of the sky because of it. But the agencies that count the debris now say some heights are crowded enough that the junk would keep growing even without new launches. Whether that stays a slow, manageable problem or turns into a costly one depends on choices being made now, about how high satellites fly, how quickly dead ones come down, and whether someone finally starts cleaning up the oldest junk.