The Zhuque-3 rocket’s stainless-steel first stage touched down on deployable legs in Gansu province on Wednesday, eight months after an earlier landing attempt ended in a fireball — a successful, if imperfect, second test that puts a private Chinese launch company in the same reusability club as SpaceX and Blue Origin for the first time.
LandSpace, a privately held Chinese launch company, successfully recovered the first-stage booster of its Zhuque-3 rocket on 19 August 2026, landing it on deployable legs at a designated site in Gansu province roughly 390 kilometres from the Dongfeng Commercial Space Innovation Pilot Zone launch site in northwestern China. The flight, designated Zhuque-3 Y2, lifted off at 7:35am Beijing time and successfully delivered its payload — the Honghu-03 satellite — to orbit via the rocket’s second stage, according to China’s state news agency Xinhua, before the recovered first stage completed a sequence of manoeuvres, including a re-entry burn, aerodynamic gliding, a final landing burn and leg deployment, to bring itself down under powered, controlled flight roughly eight minutes after liftoff.
A qualified but genuine milestone
The achievement puts LandSpace, according to multiple outlets’ framing, alongside SpaceX and Blue Origin as among a small handful of companies worldwide to have landed an orbital-class rocket booster — though the framing requires two important qualifications that responsible reporting should retain. First, this was not China’s first successful recovery of an orbital-class rocket stage in absolute terms: in July 2026, the state-backed China Academy of Launch Vehicle Technology (CALT) successfully recovered the first stage of its Long March 10B rocket using a net-capture system on a floating sea platform, a different recovery technique from LandSpace’s leg-based land landing. Wednesday’s flight was, however, China’s first successful recovery of an orbital booster using deployable landing legs, and its first successful booster recovery on land specifically — and it is the first time a privately held, rather than state-owned, Chinese company has achieved either. Second, video captured after landing showed a fire breaking out in the aft section of the recovered booster, a detail some outlets reported prominently and others omitted; LandSpace’s own statement characterised the mission overall as a full success, and the fire does not appear, on current reporting, to have compromised the primary recovery and landing objectives, though it underscores that the technique remains at an early, imperfect stage of maturity rather than a fully solved engineering problem.
The Zhuque-3’s own short flight history illustrates that immaturity candidly. The rocket’s maiden flight, in December 2025, successfully placed its payload in orbit but ended in a landing failure: Chinese state media reported an “abnormal combustion event” that prevented the intended soft landing, and video of the incident showed the booster’s first stage crashing and exploding after apparently losing an engine during its landing burn. LandSpace described that first flight as nonetheless having achieved its main technical verification objectives — engine throttling, attitude control, and validation of the broader recovery system architecture — and used the resulting flight data to refine the vehicle ahead of Wednesday’s successful second attempt, a genuinely rapid iteration cycle by the standards of orbital rocketry, where flight cadences are typically far slower than in most other engineering disciplines.
The rocket itself
Zhuque-3 is a two-stage, liquid-oxygen-and-methane-fuelled (“methalox”) rocket with a stainless-steel airframe — design choices that closely parallel, and were evidently influenced by, SpaceX’s approach with both Falcon 9 and Starship. Reported dimensions vary slightly between sources, with the rocket’s overall length given as either 66.1 or 76.6 metres depending on the outlet and possibly the specific measurement convention used; on the higher figure, it would stand taller than SpaceX’s 70-metre Falcon 9. The first stage is powered by nine TQ-12A engines, whose thrust can reportedly be throttled between 40 and 110 per cent of rated output, giving the vehicle the fine-grained thrust control that a controlled landing burn requires. LandSpace has stated the booster is designed to be reusable up to 20 times and that the rocket can carry approximately 12,500 kilograms to orbit in reusable configuration, rising to roughly 21,300 kilograms in a fully expendable flight mode that forgoes recovery. LandSpace has described Wednesday’s flight as marking the vehicle’s transition “from the technological demonstration and verification phase toward the engineering application phase of reusability” — company language that should be read as an aspiration rather than a settled fact, given that the booster has not yet actually been reflown.
Why it matters
Reusable first-stage boosters matter economically and strategically because they are, by a wide margin, the single largest cost driver in orbital launch: recovering and reflying a booster, rather than discarding it after a single use, is the innovation that allowed SpaceX to reduce launch costs sharply enough to enable the Starlink satellite constellation’s rapid buildout, fundamentally reshaping the economics of the global launch industry over the past decade. China’s ambitions in low-Earth-orbit satellite constellations — intended, in part, to compete with Starlink — depend on achieving comparable launch economics, and a demonstrated, land-recoverable, privately operated reusable booster is a necessary, if not yet sufficient, step toward that goal. That LandSpace, rather than a state-backed entity, achieved this milestone is also notable for what it signals about the maturity of China’s commercial space sector specifically: unlike CALT’s July sea-platform recovery, which drew on the resources and institutional backing of China’s primary state launch-vehicle developer, LandSpace’s achievement is a private company’s, developed with less direct state resourcing, which is a more directly comparable data point to SpaceX’s and Blue Origin’s own private-sector achievements.
The caveats that responsible coverage should retain going forward are precisely the ones the flight itself surfaced: the booster has yet to be reflown, meaning the harder and, historically, more failure-prone milestone — genuine, rapid, low-cost reusability rather than a single successful recovery — remains ahead. A post-landing fire, even if it did not compromise the mission’s stated objectives, is the kind of detail that bears watching in subsequent flights as an indicator of how close the vehicle’s recovery hardware is to true operational readiness rather than proof-of-concept. Wednesday’s flight is best read, in other words, as a significant and genuine narrowing of the gap between China’s commercial launch sector and its American counterparts — not as evidence that the gap has closed.
Sai Chaitnya Puligadda
## Key facts
– LandSpace’s Zhuque-3 Y2 rocket launched 19 August 2026 (7:35am Beijing time) from Dongfeng Commercial Space Innovation Pilot Zone; delivered the Honghu-03 satellite to orbit
– First stage landed on deployable legs in Gansu province, ~390km from the launch site, roughly 8 minutes after liftoff — China’s first land-based, leg-landed orbital booster recovery, and the first by a private Chinese company
– A separate, state-backed recovery (CALT’s Long March 10B, via sea-based net capture) occurred in July 2026, making Wednesday’s flight China’s second orbital-booster recovery overall but its first by a private firm and its first on land with legs
– A post-landing fire in the booster’s aft section was reported by multiple outlets; did not appear to compromise the mission’s stated core objectives per LandSpace’s own account
– Zhuque-3’s first flight (December 2025) reached orbit but suffered a landing failure and post-landing explosion; the design uses a stainless-steel airframe and liquid-oxygen/methane propulsion
