On Saturday afternoon, a 52-metre spacecraft fell out of orbit at 26,000 kilometres per hour, flew across Mexico and Texas glowing red-hot, flipped itself upright in mid-air, and was plucked out of the sky by two giant mechanical arms. On SpaceX's live broadcast, host Dan Huot could barely believe it: "Can you believe it? Mechazilla has caught the Starship."
It was the first time any spacecraft returning from orbit has ever been caught this way. SpaceX had already caught its Super Heavy booster three times, but the upper stage — the part actually called Starship — had always ended in the ocean. Every one of the 14 earlier Starships finished its flight in the sea. This time, it came home.
That single catch does not make Starship a finished, reusable rocket. The booster that was supposed to be caught first on the same flight failed and exploded over the Gulf of Mexico. But the ship catch is the hardest half of the problem SpaceX set itself. Here is what happened, why it was so difficult, and what still has to be proven.
What happened on Saturday
The flight began on Friday, November 13, at 4:47 p.m. Central Time, when Starship lifted off from Pad 2 at Starbase, Texas, on its fifteenth test flight. All 33 Raptor engines on the Super Heavy booster fired, and about two and a half minutes later the two stages separated in a manoeuvre called hot-staging, with Starship's own six Raptor engines igniting while still attached to the booster.
Within half an hour, Starship had fired a single Raptor engine for its orbital insertion burn and was in orbit — only the second time a Starship has reached orbit, after the first in September. Over the next 40 minutes it deployed 26 Starlink V3 satellites, SpaceX's newest internet satellites. SpaceX reported contact with every satellite soon after deployment.
Then came the long wait. Starship coasted around Earth for more than 20 hours, completing 14 orbits, while flight controllers monitored its systems. Only when everything needed for re-entry, the landing burn and the tower approach checked out did they commit to bringing it home.
The return was the part nobody had ever done. Starship relit a single Raptor engine for a deorbit burn, dropped out of orbit, and plunged through the atmosphere, glowing red as it flew over Mexico's Baja Peninsula, then Mexico and Texas, steering with its four flaps toward the coast it had left 22 and a half hours earlier. On a grey, cloudy day at Starbase, the ship burst through the clouds, fired three Raptor engines, flipped vertical, and sank into the tower's waiting arms at about 3:20 p.m. Central Time on Saturday, November 14.
The flight in numbers
- Launch: Friday, November 13, 4:47 p.m. CST, Pad 2, Starbase, Texas
- Catch: Saturday, November 14, about 3:20 p.m. CST — 22.5 hours later
- Orbits: 14, at 275 kilometres (170 miles)
- Satellites: 26 Starlink V3, all contacted
- Ship height: 52 metres (171 feet); full stack: 124 metres (407 feet)
- Re-entry speed: about 26,000 km/h (16,000 mph)
- Earlier booster catches: 3 — this was the first ship catch
Why catching the ship is harder than catching the booster
It is tempting to think that once you have caught one part of a rocket, catching the other is the same trick twice. It is not. The Super Heavy booster never reaches orbit: it separates minutes after launch and comes back at a few thousand kilometres per hour, a problem of steering and timing. The ship is a different challenge — speed and heat.
Starship comes back from orbit, arriving at roughly 26,000 kilometres per hour and shedding almost all of that speed by dragging itself through the atmosphere, where the air ahead turns to plasma. Its heat shield — thousands of black hexagonal tiles covering one side of the ship — must survive temperatures that would melt ordinary metal. Only after the ship has slowed to a crawl, flipped from belly-first to tail-first, and relit its engines can the tower arms try to grab it.
That is why the long coast mattered. SpaceX spent more than 20 hours checking the ship's health before committing to the return — a damaged heat shield or a sick engine would have meant a planned splashdown instead. The decision to attempt the catch was only made about two hours before the return.
Then there is precision. The tower arms close on two small catch pins on the side of the ship. It has to arrive in exactly the right place, at exactly the right speed, with almost no sideways drift. Get it right and the ship simply hangs there, ready to be lowered, inspected, refuelled and — one day — flown again.
The booster that did not come home
For all the celebration, the flight was only half a success. The plan was a "double catch": both the booster and the ship returning to the same tower on the same flight, something no orbital rocket system has ever achieved. That part failed within minutes of launch.
After separation, the booster relit 32 of its planned 33 Raptor engines for the boostback burn that sends it back toward land. But the burn ended early because of low fuel mass, according to SpaceX, and during the coast the booster's guidance, navigation and control system detected a problem and triggered an automated abort. The booster was diverted offshore in the Gulf of Mexico, carried out a landing burn, hovered over the water, and then its flight termination system fired. It exploded just above the sea.
This was not a random accident. The abort system did what it was designed to do: when the booster could not be sure of reaching the tower safely, it took itself out to sea rather than risk the launch site. SpaceX says the root cause of the low fuel mass is still under investigation. Note that this was the first catch attempt with the upgraded Super Heavy V3 booster, a taller 72-metre version built for Moon and Mars — new hardware, new failure modes.
What Starship could make possible
Why does any of this matter beyond the spectacle? Because a rocket you can catch, refuel and fly again changes the economics of everything that goes to space.
Today, almost every orbital rocket is thrown away after one flight. SpaceX already reuses Falcon 9 boosters, which land on legs, and that alone cut the cost of reaching orbit dramatically. Starship goes further: no landing legs, no ocean recovery ships, no weeks of refurbishment. In theory, a caught Starship can be stacked on a caught booster, refuelled, loaded, and launched again quickly — like an aeroplane turning around between flights rather than being rebuilt after each one.
The most immediate use is Starlink. This flight carried 26 of the new, larger V3 satellites, and SpaceX wants to launch thousands of them. A cheap, rapidly reusable rocket is what makes a constellation of that size affordable. NPR reported that SpaceX spent more than $1 billion on Starship in the first nine months of 2026 alone.
The bigger uses are further away. NASA plans to use a version of Starship as the lander that will carry astronauts to the Moon under the Artemis programme, and SpaceX's long-term goal is Mars. Neither is possible without a vehicle that can be refuelled in orbit, flown many times, and trusted with a crew — none of which has been demonstrated yet.
Cheap heavy lift would also change space science. The great observatories of recent years — from the Euclid telescope's first discoveries to the James Webb Space Telescope's lopsided exoplanet atmosphere — were all shaped by the size and weight limits of the rockets that carried them. A rocket lifting 100 tonnes or more at a fraction of today's cost would let astronomers build bigger telescopes, and fly more of them. Even the strangest finds, like the most distant fast radio burst Webb has found, hint at how much more there is to see once getting into space stops being the hardest part.
What still has to be proven
A single catch is a proof of concept, not a production line. Here is what remains.
Fly a caught ship again. The point of catching Starship is to reuse it. Nobody has yet reflown a caught Starship, or even a caught booster, on a second mission. Until a recovered vehicle flies again — and then again — reusability is a promise, not a fact. The first job now is inspecting this ship: its heat shield tiles, engines, and structure, to find out what 22 hours in space and a fiery re-entry did to it.
Catch both stages on one flight. The double catch is still elusive. SpaceX must find why the booster's boostback burn ended early, fix it, and do both catches on the same day. Until then, every Starship flight still throws away half the rocket.
Refuel in orbit. A Starship going to the Moon or Mars cannot carry enough fuel for the whole trip at launch. It must be refuelled by tanker Starships in orbit — a manoeuvre never attempted at this size, and arguably a bigger challenge than the catch itself.
Prove it safe for people. Cargo and satellites are one thing; astronauts are another. A crewed Starship needs escape systems, life support proven over long flights, and a safety record NASA and regulators will accept. That work is years away.
Do it fast, and do it often. Rapid reusability means a turnaround measured in days, not months. SpaceX's whole business case — Starlink, lunar landings, Mars — depends on flying Starships frequently and cheaply.
What happens next
SpaceX has not announced a date for the next flight, but the programme rarely waits long. The priorities are clear: investigate the booster's low fuel mass and guidance abort, inspect the caught ship piece by piece, and try the double catch again. The company also plans to start launching Starships from Florida within the coming months.
For anyone watching, the thing to look for next time is not whether the ship can be caught again — it probably can — but whether the booster comes home too, and whether either vehicle flies a second time. Those tests turn a spectacular catch into a working transport system.
Saturday's catch was historic: the first time a spacecraft has come back from orbit and been caught by the tower it launched from. It was also only half the job — the other half is at the bottom of the Gulf, waiting for an explanation.