China has landed two orbital-class rocket boosters, but neither has flown again. The landings came five weeks apart in 2026 and used very different methods. On 10 July, the state-built Long March 10B was caught in a net on a sea platform. Then, on 18 August, LandSpace’s Zhuque-3 touched down on four legs in Gansu province.
Those two results put China in a group of two countries that have brought an orbital first stage home, alongside the United States, where SpaceX, Blue Origin, and Rocket Lab all recover boosters.
But recovery is only half the story: no Chinese booster has yet flown a second time. So every program below is measured against the same two questions: Has a stage come back? And has one gone up again?
Every Chinese Reusable Launch Program
| Vehicle | Builder | Propellant | Recovery method | Status as of September 2026 |
|---|---|---|---|---|
| Long March 10B | CASC and CALT, state-owned | Kerosene and liquid oxygen | Net catch on a sea platform | First stage recovered 10 July 2026. CASC aims to refly by end of 2026 |
| Zhuque-3 | LandSpace | Methane and liquid oxygen | Landing legs, downrange pad | Landed 18 August 2026. Booster toppled after a post-landing fire |
| Tianlong-3 | Space Pioneer | Kerosene and liquid oxygen | Landing legs planned | First orbital flight failed shortly after launch, April 2026 |
| Hyperbola-3 | iSpace | Methane and liquid oxygen | Legs, with sea recovery under study | In development. No orbital recovery attempt yet |
| Pallas-1 | Galactic Energy | Kerosene and liquid oxygen | Vertical landing planned, not attempted on the debut flight | Reached orbit on an expendable debut flight 1 September 2026. Recovery attempts are planned on later flights |
| Nebula-1 | Deep Blue Aerospace | Kerosene and liquid oxygen | Landing legs | High-altitude test failed near touchdown. Program continuing |
| Long March 12A | SAST, state-owned | Kerosene and liquid oxygen | Vertical landing planned | Reported close to an orbital test flight |
The Long March 10B Came Home in a Net
The Long March 10B first stage came home in a net on a sea platform rather than on legs. That distinction is easy to lose in headlines and it changes the engineering. A caught stage carries no landing legs, so it saves that mass and converts it into payload. It also has almost no margin for error, because the recovery hardware sits in one fixed place and the booster has to arrive at it.
The flight lifted off from Hainan on 10 July 2026 and made China the second country to recover an orbital-class booster, according to SpaceNews. The China Aerospace Science and Technology Corporation, or CASC, said it intends to fly that stage again before the end of 2026.
The 10B is the reusable member of the Long March 10 family, the rockets built for China’s crewed lunar program. Its variants are covered on the Long March 10 page.
Zhuque-3 Landed on Legs and Then Fell Over
Zhuque-3 completed the first leg landing by a Chinese orbital booster and lost the stage afterward. The rocket flew from the Dongfeng commercial zone at Jiuquan on 18 August 2026, delivered the Honghu-03 satellite, and put its first stage down upright about 390 kilometers downrange in Minqin County. Some time later a fire during propellant offloading weakened a leg, and the booster tipped.
Guidance, engine relight and touchdown all worked. Ground handling, the far less glamorous half of reuse, did not. Full details are on the Zhuque-3 page.
The Private Field Behind LandSpace
Four private Chinese companies are building reusable vehicles and none has recovered an orbital stage yet.
- Space Pioneer flies Tianlong-3, a kerosene rocket in the Falcon 9 size class. Its first orbital attempt failed shortly after launch in April 2026, before any recovery could be attempted.
- iSpace is developing Hyperbola-3, a methane vehicle with vertical landing, and has studied recovering it downrange at sea.
- Galactic Energy has a working revenue business in the small expendable Ceres-1. It reached orbit with the larger Pallas-1 on 1 September 2026, an expendable debut flight from the Dongfeng commercial zone at Jiuquan that carried no customer payload, according to SpaceNews. Future flights are set to attempt first-stage recovery. Selling Ceres-1 launches funds the reusable program.
- Deep Blue Aerospace flew a high-altitude vertical landing test of Nebula-1 that failed close to touchdown, which is the hardest and most instructive part of the profile to get wrong.
The State and Private Programs Answer to Different Customers
China is running two reuse programs with different owners and different incentives. CASC builds the Long March vehicles for national missions, so its reusable rocket has to serve crewed lunar flights, constellation deployment, and military payloads on a schedule set by government planning. Cost per kilogram matters to it, but availability matters more.
The private firms answer to investors and to constellation operators buying launches by the ton. LandSpace filed for a listing on the Shanghai Stock Exchange’s STAR Market at the end of 2025 seeking about 7.5 billion yuan, and its prospectus projects profitability in 2029. A company with that timeline needs a booster that flies often and cheaply. A five-year state production plan does not produce one.
The two tracks also pick different hardware. CASC chose a sea catch, LandSpace chose legs. Neither has declared the other wrong, and China is effectively testing both approaches in parallel at national expense.
Why China Is Chasing Reuse
The demand comes from two approved broadband constellations. Guowang and Qianfan together aim for more than 20,000 satellites. Filling them at the launch rate China ran through the early 2020s would take decades, because each expendable Long March flight carries a modest batch and then is gone.
That arithmetic is the reason a state program with no commercial pressure started building landing legs. China flew a national record 92 orbital launches in 2025 and has targeted roughly 140 for 2026, according to SpaceNews. Even that rate leaves the constellations short. Reuse is the only lever that raises flight count without building a proportionate number of new rockets.
Satellite counts and deployment schedules for both networks are on the China satellite constellations page.
Turning a Recovered Booster Around Takes Months
The step China has not publicly cleared is turning a recovered booster around. SpaceX needed roughly a year and a half between its first Falcon 9 landing in December 2015 and its first reflight in March 2017, and that was with a company already flying dozens of missions a year. The work in that gap is inspection, engine teardown, structural checks, and requalification, and none of it makes the news.
Two Chinese programs are now inside that gap at the same time. CASC has stated an end-of-2026 target for the Long March 10B and LandSpace a six-month target for Zhuque-3. Both dates were set before the Zhuque-3 booster tipped over.
What Would Change the Assessment
One event settles it: a Chinese booster launching a payload to orbit for the second time. That would turn LandSpace’s 20,000 yuan per kilogram target from a projection into a measured rate. A second change would come from a repeat performance. If Zhuque-3 or the Long March 10B lands cleanly two or three more times, the 2026 results read as a procedure rather than a pair of good days.
A third would come from Space Pioneer or iSpace reaching orbit and recovering a stage, because a second private company clearing the same bar would show the capability sits in the sector rather than in one firm.
For the numbers behind the comparison people usually want next, see how China’s reusable programs stack up against SpaceX, or read the mechanics of how reusable rockets work on the reusable rockets guide.