In July 2026, one Falcon 9 booster flew for the 36th time. One day later, China recovered an orbital-class rocket stage for the first time. These two events showed the beginning and the current edge of the same technological shift.

SpaceX spent years turning rocket landing from a dramatic experiment into a normal part of launch operations. Blue Origin developed a parallel path with New Shepard and later landed the much larger New Glenn booster. China then recovered Long March 10B on its first orbital flight.

The later vehicles did not have an easy job. They still needed engines, structures, software, ships, and flight tests that worked together. However, they entered a world in which Falcon 9 had already shown that orbital booster recovery was possible and commercially useful.

Falcon 9 proved that a booster could return.
Reflight proved that the same hardware could earn money again.
High launch cadence turned reuse into an industrial system.

Powered landing moved from short test hops to repeated orbital-booster reuse.

What Is Falcon 9?

Falcon 9 is a two-stage rocket built by SpaceX. It carries satellites, cargo spacecraft, and crewed Dragon vehicles into orbit. The rocket first reached orbit on June 4, 2010.

The current vehicle is about 70 meters tall and 3.7 meters wide. Its first stage uses nine Merlin engines. The second stage uses one Merlin Vacuum engine. Both stages burn liquid oxygen and rocket-grade kerosene, also called RP-1.

SpaceX lists a maximum payload of 22.8 metric tons to low Earth orbit. That public maximum should not be treated as the payload of every reusable mission. A returning first stage must keep propellant for its descent, so recovery reduces the payload available for some trajectories.

Falcon 9 Element	Role
Nine first-stage engines	Provide liftoff thrust and engine-out redundancy during part of ascent.
Second stage	Continues toward orbit after the first stage separates.
Grid fins	Steer the returning booster through the atmosphere.
Restartable engines	Slow the stage during entry and final landing.
Landing legs	Deploy shortly before touchdown and support the booster after landing.

The first stage performs most of the work near Earth, but it does not continue to orbit. SpaceX saw this large and expensive stage as the best place to begin reuse.

How Does a Falcon 9 Booster Return?

After stage separation, the upper stage keeps accelerating toward orbit. The first stage turns around and prepares to fall back through the atmosphere.

For a return to the launch site, the booster performs a boostback burn that changes its horizontal motion. For a droneship landing, the ship waits farther downrange, so the booster can save some return propellant.

The stage then performs an entry burn to reduce speed and heating. Grid fins guide it through the denser atmosphere. Near the surface, one or more engines ignite for the landing burn. The legs deploy, and the vehicle touches down vertically.

A Falcon 9 first stage lands at Landing Zone 1 on February 19, 2017. Photo: SpaceX. CC0 public-domain dedication via Wikimedia Commons.

Every part of this sequence must work at the right time. The booster must know its position, speed, angle, remaining propellant, engine condition, and landing target. A small error early in the descent can become a large miss near the ground.

The Grasshopper Years: Learning Close to the Ground

SpaceX did not begin by trying to land a booster after an orbital mission. It first built Grasshopper, a vertical takeoff and vertical landing test vehicle based on a Falcon 9 first-stage tank and one Merlin engine.

Grasshopper began short flight tests in 2012 at SpaceX’s McGregor facility in Texas. The early hops looked modest, but they tested the core control loop: rise, hold position, move sideways, descend, and land on the same legs.

This approach reduced risk. Engineers could run many tests near the ground, recover the vehicle immediately, inspect it, and change the software. Grasshopper did not reproduce the speed and heating of an orbital mission, but it taught SpaceX how to control a tall rocket during powered descent.

The development method mattered as much as the vehicle. SpaceX separated the problem into smaller steps, learned from each flight, and then moved the same control ideas into full-scale missions.

From Ocean Tests to a Real Landing Pad

SpaceX began full-scale first-stage return experiments in September 2013. That Falcon 9 stage reentered the atmosphere, but it did not complete a controlled recovery.

In April 2014, during the CRS-3 cargo mission, a Falcon 9 first stage achieved a controlled vertical touchdown in the Atlantic Ocean. The stage was not recovered intact, but the flight showed that its engines, guidance, and landing sequence could reduce the vehicle to near-zero vertical speed at the surface.

Ocean tests gave SpaceX room to fail without hitting a ship or landing zone. They also allowed the company to collect full-scale flight data while the main payload mission continued normally.

The next step was harder. A landing platform offers little room for error, and an offshore ship moves with waves and wind.

How Many Landing Attempts Did SpaceX Need?

The answer depends on what we count. Grasshopper hops were prototype tests. Ocean touchdowns tested the return sequence but did not recover an intact booster. Platform landings were the first attempts to bring the stage back as usable hardware.

Using that narrower platform-landing definition, SpaceX succeeded on its third attempt.

Attempt	Mission and Date	Result
1	CRS-5 — January 10, 2015	Reached the droneship but crashed after the grid-fin hydraulic system ran out of fluid.
2	CRS-6 — April 14, 2015	Reached the deck and touched down, but excessive sideways motion caused the stage to tip over.
3	ORBCOMM-2 — December 21, 2015	Landed upright at Landing Zone 1, completing the first recovery of an orbital-class rocket stage.

This “third attempt” does not mean SpaceX solved reuse in three flights. Several years of prototype hops, ocean tests, engine work, software development, and earlier return experiments came first.

The failures were also useful. The first droneship attempt exposed a hydraulic-fluid limit. The second showed how engine response and lateral motion could destroy an otherwise accurate approach. Each failure turned a broad problem into a specific engineering change.

Landing Was Only the First Half

The December 2015 landing proved recovery on land. SpaceX still needed to master offshore recovery and prove that a recovered booster could fly again.

A January 2016 stage reached a droneship but tipped over when one landing leg failed to lock. On April 8, 2016, the CRS-8 booster landed successfully on the droneship Of Course I Still Love You. It was SpaceX’s fourth serious droneship landing attempt.

SpaceX then inspected that CRS-8 booster, prepared it again, and launched it on the SES-10 mission on March 30, 2017. The same first stage landed for a second time after the mission.

This was the more important business milestone. A recovered rocket is an engineering success. A reflown rocket becomes productive hardware.

By February 2025, SpaceX reported more than 384 Falcon first-stage reflights. By July 2026, one booster had completed 36 missions, while the Falcon fleet had passed 600 booster landings. The question was no longer whether a booster could return. It was how many flights it could complete and how quickly it could fly again.

Why Falcon 9 Became the Industry Model

Falcon 9 did not invent vertical rocket landing. Earlier experimental vehicles had already demonstrated parts of the idea, and Blue Origin developed its own vertical landing program in parallel.

Falcon 9 changed the industry because it connected recovery to real orbital missions. It carried commercial satellites, NASA cargo, astronauts, national-security payloads, and SpaceX’s own Starlink satellites while regularly bringing the first stage home.

That record revealed a reusable design language that other programs could study:

1. Use restartable, throttleable liquid engines.
2. Control the stage with aerodynamic surfaces during descent.
3. Land vertically to avoid wings and runways.
4. Use offshore platforms when returning to the launch site would consume too much propellant.
5. Collect flight data, inspect recovered hardware, and improve the next vehicle.
6. Create enough launch demand to keep the reusable fleet busy.

The last point is easy to miss. Reuse does not lower cost by itself. A company needs frequent missions, rapid processing, reliable engines, available launch pads, recovery ships, and customers. Starlink gave SpaceX a large internal demand that helped the company use its boosters repeatedly.

Jeff Bezos: New Shepard First, Then New Glenn

Blue Origin reached an important milestone before Falcon 9’s first intact landing. On November 23, 2015, New Shepard flew above 100 kilometers and landed its booster vertically in West Texas.

On January 22, 2016, Blue Origin flew and landed the same booster again. This was the first successful reuse of a vertically landed rocket booster.

However, New Shepard is suborbital. It rises almost vertically, separates from its capsule, and returns near its launch site. Falcon 9 sends an upper stage and payload toward orbit, while its first stage returns from a much faster and more complex downrange trajectory.

Blue Origin later moved the same basic philosophy into New Glenn, a 98-meter orbital launch vehicle. Its first flight reached orbit on January 16, 2025, but the booster was lost during descent. On its second mission, November 13, 2025, New Glenn placed NASA’s ESCAPADE spacecraft into the planned orbit and landed its first stage on the offshore platform Jacklyn.

New Glenn therefore achieved an orbital-class booster landing on its second flight. That was much faster in public flight count than Falcon 9’s early program. Yet Blue Origin had already built a decade of landing experience through New Shepard, and it could study the operational lessons that Falcon 9 had made visible.

Long March 10B: Recovery on the Maiden Flight

China moved one step faster in public flight count. Long March 10B reached orbit and recovered its first stage on its maiden mission on July 10, 2026.

The Chinese vehicle did not copy every Falcon 9 feature. Instead of carrying large landing legs, its first stage used four hooks to engage a net system on an offshore platform. The design moved more recovery hardware onto the ship.

The first-flight success shows how quickly a later program can integrate proven principles. Engineers no longer need to ask whether a tall orbital booster can restart its engines, steer through the atmosphere, and reach a small offshore target. Falcon 9 and New Glenn have already answered that question.

They still need to answer another question: can their own engines, structures, software, and recovery system do it reliably?

Program	First Relevant Milestone	First Orbital-Class Recovery	Reuse Status by July 2026
Falcon 9	Grasshopper tests began in 2012; full-scale return experiment in 2013	December 2015, third platform-landing attempt	Routine reuse; one booster reached 36 flights
New Glenn	New Shepard landed in 2015 and was reused in 2016	November 2025, second New Glenn orbital flight	Recovery demonstrated; repeated New Glenn reuse not yet established
Long March 10B	Years of Chinese ground, engine, and vertical-landing development	July 2026, maiden orbital flight	Recovery demonstrated; first reflight still to be proven

The comparison needs caution. Public flight count is not the same as total development time. SpaceX, Blue Origin, and China used different test programs, vehicles, mission profiles, and levels of public disclosure.

Still, the pattern is clear. The pioneer paid the highest learning cost. Later programs could watch the failures, study the architecture, improve simulation, and begin full-scale flight with fewer unknowns.

The Hidden Lesson: A Technology Becomes a Template

A successful technology changes more than one company. It changes what competitors believe is possible, what investors will fund, what suppliers will build, and what governments will request.

Before Falcon 9, reusable launch vehicles were often linked to large spaceplane programs or experimental demonstrators. Falcon 9 showed another path: recover the most valuable stage, keep the upper stage simple, and improve the system through frequent commercial flights.

New Glenn adopted a reusable orbital first stage with offshore landing. Long March 10B used the same broad operational logic but replaced legs with a net-capture system. Both vehicles show how an industry learns: copy the principle, change the implementation, and try to remove the pioneer’s weaknesses.

The fastest follower may reach the first milestone quickly. The harder test is what follows: reflight, turnaround, maintenance, reliability, and launch cadence.

What to Watch Next

1. Long March 10B reflight: Will China launch the same recovered stage before the end of 2026?
2. New Glenn repeatability: Can Blue Origin land and refly orbital boosters across several missions?
3. Inspection work: How much hardware must each program replace after landing?
4. Turnaround time: Does reuse take months, weeks, or days?
5. Flight demand: Do satellite constellations and government missions provide enough launches to make reuse economical?

Conclusion

Falcon 9 began as a conventional two-stage rocket, but SpaceX gradually added a second mission to every launch: bring the first stage home. Grasshopper tested the control system. Ocean touchdowns tested the full-scale descent. Failed platform landings exposed specific weaknesses. The 2015 landing proved recovery, and the 2017 SES-10 mission proved orbital-class reflight.

Blue Origin showed that vertical landing could also support a reusable suborbital system, then reached an orbital landing with New Glenn. China reached its first Long March 10B recovery on the vehicle’s maiden flight.

That speed is evidence of technological learning. Falcon 9 did more than solve a hard problem. It made the solution visible enough for the rest of the world to build from it.

Key Vocabulary & Phrases

Powered descent (noun)
A descent in which engines provide thrust to control speed and direction.
Example: Grasshopper helped SpaceX test powered descent close to the ground.

Flight-proven (adjective)
Already tested during a real flight rather than only on the ground.
Example: SES-10 launched on a flight-proven Falcon 9 first stage.

Downrange (adverb or adjective)
Farther along the path of a rocket after launch.
Example: A droneship waits downrange so the booster uses less return propellant.

Iteration (noun)
One cycle of testing, learning, and improving a design.
Example: Each failed landing gave SpaceX information for the next iteration.

Operational model (noun)
A practical system for performing the same activity repeatedly.
Example: Falcon 9 turned rocket landing into an operational model.

Technology transfer (noun)
The movement of knowledge or methods from one project, company, or country to another.
Example: Public demonstrations can speed technology transfer without direct copying.

Next in This Series

• Reusable Rockets Explained: Why China’s Long March 10B Recovery Matters
• Methane Engines, 3D Printing, and the New Rocket Design Language
• The Economics of Reuse: When Does Recovering a Rocket Save Money?
• Where Korea Stands in the Reusable Launch Race

References

• Falcon 9 Vehicle Overview and Specifications — SpaceX.
• Falcon 9 Demo Mission — SpaceX.
• The Annual Compendium of Commercial Space Transportation: 2012 — U.S. Federal Aviation Administration.
• SpaceX CRS-6 Mission Press Kit — NASA and SpaceX.
• ORBCOMM-2 Mission — SpaceX.
• SES-10 Mission — SpaceX.
• Falcon Payload User’s Guide — SpaceX.
• Blue Origin Makes Historic Rocket Landing — Blue Origin.
• Launch. Land. Repeat. — Blue Origin.
• Blue Origin’s New Glenn Reaches Orbit — Blue Origin.
• New Glenn Mission NG-2 — Blue Origin.
• China Achieves First Controlled Rocket Recovery — China Aerospace Science and Technology Corporation.
• Falcon 9 First Stage Lands on LZ-1 — SpaceX image, CC0 via Wikimedia Commons.