SpaceX Reusable Rockets: How They Work and Why They Changed Everything

On December 21, 2015, a 14-story rocket booster streaked down from the Florida sky, slowed itself with a last-second engine burn, and landed upright on four legs at Cape Canaveral — roughly where it had lifted off from nine minutes earlier. The crowd watching erupted. Commentators who had been covering rocket launches for decades couldn't find words. SpaceX engineers in the control room screamed and hugged each other.

That landing changed the economics of space travel. For the entire history of the rocket age, rocket boosters — the most expensive part of any launch vehicle — were thrown away after a single use, falling into the ocean. SpaceX proved they didn't have to be. The Falcon 9 booster that landed that night in 2015 was eventually reflown. Some Falcon 9 boosters have now flown more than 20 times.

Key takeaways

• SpaceX Falcon 9 boosters land themselves using a combination of grid fins, cold gas thrusters, and a propulsive burn of the main engines
• Reusability has reduced SpaceX's cost per kilogram to orbit by roughly 80% compared to expendable rockets
• The Falcon 9 first stage contains 9 Merlin engines; it uses 3 (sometimes 1) during the landing burn
• Drone ships named "Of Course I Still Love You" and "A Shortfall of Gravitas" catch boosters landing at sea
• As of 2026, individual Falcon 9 boosters have been successfully reflown more than 20 times
• Starship, SpaceX's next-generation vehicle, uses a "mechazilla" catch system to catch the booster with robotic arms

The problem: Rockets are expensive because they're disposable

Before SpaceX, the economics of launch were simple and brutal. A rocket cost anywhere from $100 million to $400 million to build. You used it once. The first stage, which contains the most powerful and expensive engines, fell away two and a half minutes after launch and crashed into the ocean. The second stage completed the mission and then either burned up on reentry or became orbital debris. Every launch required building an entirely new rocket.

This made the cost of reaching orbit enormous. The Space Shuttle was designed to be reusable but the refurbishment costs were so high that it ended up costing more per flight than a new expendable rocket. The solid rocket boosters were recovered and reused, but the external tank was lost every flight.

SpaceX's founder Elon Musk famously compared the situation to aviation: imagine if you threw away a Boeing 747 after every flight. Airfare would cost a million dollars. The only way to make space accessible — for science, commerce, or eventual human settlement of other planets — was to make rockets reusable like aircraft.

The engineering: How do you land a rocket?

Landing a used rocket booster is one of the most technically demanding maneuvers in modern engineering. The Falcon 9 first stage is about 14 stories tall and weighs about 25,000 kg (empty). After separating from the upper stage at an altitude of about 80 km (50 miles) and a speed of about 5,500 mph, it needs to:

1. Flip around to point engines downward
2. Decelerate enough to reenter the atmosphere without burning up
3. Deploy four grid fins for aerodynamic control during atmospheric descent
4. Navigate to a precise landing target (either the launch pad or a drone ship at sea)
5. Extend landing legs
6. Fire engines one final time to slow from about 150 mph to near-zero, landing with centimeter precision

Each step requires the onboard flight computer to make thousands of decisions per second. The "entry burn" uses three of the nine Merlin engines to slow the booster before it hits the thickest part of the atmosphere, protecting the engine bay from heat. Grid fins — four large metal fins that fold out after stage separation — provide aerodynamic control during the hypersonic descent, pointing the booster like a bullet toward the target. The "landing burn" — sometimes using just a single engine — brings the booster to a hover just above the landing pad, then cuts it to let the legs absorb the final touchdown.

The whole sequence from stage separation to landing takes about eight to nine minutes. The booster has no pilot. It is entirely autonomous, guided by GPS, inertial navigation, and onboard cameras that detect the landing pad's position.

Drone ships: Landing on a floating platform at sea

When a Falcon 9 is launching a heavy payload to a high orbit, it burns more propellant reaching orbit and doesn't have enough fuel left for a "return to launch site" landing. In these cases, the booster lands on one of SpaceX's autonomous drone ships — floating ocean platforms that can be precisely positioned downrange under the flight path.

The drone ships are converted ocean barges equipped with four diesel-powered azimuth thrusters that hold them in position. They are named after spacecraft in the Culture novel series by Iain M. Banks: "Of Course I Still Love You" operates in the Atlantic, "A Shortfall of Gravitas" replaced the earlier "Just Read the Instructions" in the Pacific. The names were chosen by SpaceX engineers who are fans of the books.

Landing on a drone ship is harder than landing on the launch site. The ship bobs in ocean swells. The booster approaches at 150+ mph from a slightly different angle on every flight. Winds may have shifted since launch. The first several attempts at drone ship landings failed spectacularly — boosters tipping over, exploding, coming in too fast. SpaceX published a blooper reel of failures that accumulated before the first successful drone ship landing in April 2016.

Economics: What reusability actually saves

A Falcon 9 rocket costs roughly $60 million to build. The propellant burned during a flight costs about $300,000 — less than 1% of the total cost. If you can reuse the rocket even a handful of times, you slash the per-flight cost dramatically. SpaceX's current launch prices — around $67 million for a standard Falcon 9 commercial launch — are significantly lower than comparable rockets from competitors.

The savings are even more dramatic when you consider that SpaceX sometimes launches the same booster within two to three weeks of a previous mission. In the early era of expendable rockets, rocket companies needed to build a new vehicle for each launch — a process taking months. SpaceX can turn around a Falcon 9 in weeks because the booster just needs inspection, refueling, and any required replacement of parts exposed to the most heat.

The competitive pressure this created has forced the entire launch industry to either lower prices or lose business. United Launch Alliance, which previously charged $400 million per launch for national security missions, has been forced to fundamentally redesign its business model. Europe's Ariane 6 was partially redesigned to be cheaper. China's launch providers lowered prices. Reusability triggered a price war that has made it cheaper than ever to reach orbit.

Starship: Taking reusability to the extreme

If Falcon 9 reusability was revolutionary, SpaceX's Starship program is existential. Starship is a fully reusable two-stage rocket system — the largest ever built — designed to be reflown within hours of landing, like an aircraft. The first stage, called Super Heavy, carries 33 Raptor engines generating 17 million pounds of thrust — more than twice the Saturn V. The upper stage, called Starship, can carry 100+ tons to low Earth orbit.

The most dramatic innovation is the catching system. Rather than using legs to land on a platform, Super Heavy is caught by two large robotic arms — nicknamed "mechazilla" — built into the launch tower. The arms reach out and grab the booster by its grid fins as it descends, eliminating the need for heavy landing legs and allowing the booster to be immediately remounted on the launch mount for rapid reflight.

Starship has been flying test missions since 2023, with increasing success. The vehicle is central to NASA's Artemis program, which has contracted SpaceX to use a modified Starship as the Human Landing System to take astronauts from lunar orbit to the Moon's surface.

FAQ

How does the computer know where to land? The Falcon 9 booster uses GPS for position and inertial navigation (accelerometers and gyroscopes) for attitude. As it approaches the landing pad, cameras detect the pad's position markers to refine the final guidance. The flight software runs a continuous real-time trajectory optimization, adjusting the landing burn timing and thrust level to account for actual conditions versus predictions.

What parts of the rocket are NOT reused? On Falcon 9, the second stage is not recovered — it's too small to carry enough propellant for reentry and landing while still delivering useful payload. The payload fairing (the nose cone that protects satellites during launch) is recovered by ships with nets, dried out, and reused. The first stage is refurbished between flights, with some parts — like the heat shield on the engine bay — inspected and sometimes replaced.

Has a Falcon 9 ever failed to land? Yes, though failures have become increasingly rare. In the early years of reusability testing (2015–2016), many landing attempts ended in crashes. Since then, the success rate has approached 99%. The most common failure mode in early tests was coming in too fast or running out of hydraulic fluid in the grid fin actuators. A handful of failures have occurred during operational missions due to hardware issues.

How long does refurbishment take between flights? SpaceX has achieved turnaround times of under two weeks on some boosters. Typical refurbishment involves inspection of all engine components, replacement of anything showing unusual wear, reloading of the nitrogen gas used for reaction control thrusters, and testing before the next launch. The goal for Starship is to achieve airline-style turnaround of hours rather than weeks.

Will all rockets become reusable? Most new rocket designs now include reusability as a requirement. Blue Origin's New Glenn, ULA's Vulcan (which plans booster engine reuse), China's Tianlong-3, and others are all designed with reusability in mind. The era of fully expendable rockets is effectively over for commercial launch providers — SpaceX's success proved that the economics make reusability the only viable path forward.

---

Read next:

• SpaceX Starship Explained — Rocket Built for Mars
• Why Rocket Launches Cost So Much — And How SpaceX Is Changing That
• How Rocket Engines Actually Work — From Kerosene to Ion Thrusters