Could the key to Mars depend on mastering one engine’s restart in space? For SpaceX, the tenth Starship flight was more than an event it was a test bed for capabilities that will shape the future of human spaceflight.
On August 26, the Starship-Super Heavy system, standing at 403 feet tall, took off from Starbase, Texas, with the thrust of 33 Raptor engines fueled by methane with a total force of around 16 million pounds. The flight was smooth with stage separation about three minutes into flight through hot staging, a method where the six Raptors in the upper stage fire before the booster separates completely. The technique, tried initially during previous flights, reduces velocity loss and conserves propellant efficiency.
The “Ship,” upper stage marked two achievements that had previously evaded missions. Initially, it successfully deployed eight dummy Starlink satellites with its so-called “pez dispenser” payload bay a system intended for quick, sequential release of satellites. This deployment was important not only for SpaceX’s commercial broadband goals but also for proving Starship could successfully manage large-scale payload operation for NASA’s Artemis lunar missions and future Mars cargo trips. Second, it conducted the second-ever in-space restart of a Raptor engine, a capability critical to deorbit burns and interplanetary trajectory adjustments. The Raptor’s full-flow staged combustion cycle, running at chamber pressures above 300 bar, renders such restarts technically demanding, particularly following extended exposure to the thermal extremes of space.
Meanwhile, the Super Heavy booster performed a series of test landing burns prior to its scheduled splashdown in the Gulf of Mexico. One of its three center engines was deliberately shut down partway through burn to mimic failure, causing the booster to use a mid-ring backup engine. The booster switched to a two-engine hover before shutting down thrust and splashing into the ocean. This test pushes the limits on engine-out landing situations, an important aspect of operational reusability.
Reentry was meant to be brutal. Engineers had removed heat shield tiles in selected regions and replaced them with substitute metallic tiles, one of which featured active cooling, to test thermal loads. Working catch fittings were structurally and thermally tested as well hardware that will ultimately enable Starship to be “caught” by the launch tower’s mechanical arms instead of land in the ocean. Although it lost segments of a protective skirt and partially melted a control flap hinge, Ship was able to maintain stability by actuation of the flaps, employing its four aerodynamic surfaces to regulate descent. It landed on a flip and soft splashdown in the Indian Ocean prior to its predicted post-landing breakup.
The success of the mission follows a series of failures earlier in 2025, such as Ship 36 being lost in a static fire test in June when a composite overwrapped pressure vessel failure occurred, and a series of in-flight disintegrations caused by pressurization system malfunctions and engine hardware issues. Hardware redesigns have since been put in place by SpaceX, including a redesigned diffuser for the main fuel tank pressurization system and lowered operating pressures for COPVs, in addition to expanded inspection procedures.
Regulatory momentum could accelerate the program’s cadence. The FAA has authorized up to 25 Starship launches annually from Texas without a new environmental review, and an August executive order from President Donald Trump rolled back certain federal oversight measures for commercial spaceflight. Elon Musk has previously suggested a target of one Starship launch every three to four weeks, though Flight 10’s delays caused by both weather and ground system issues show the challenges of achieving such frequency.
For NASA’s Artemis III mission, planned for no sooner than 2027, Starship will be the Human Landing System, needing 10 to 20 in-orbit tanker missions to refuel before departing for lunar orbit. This will need to be accomplished with quick launch turnaround, accurate cryogenic propellant transfer, and prevention of boil-off over long periods ability not yet proven at scale. The same technologies will underpin Mars missions, where methane and liquid oxygen could be produced on-site via the Sabatier reaction, enabling return trips without Earth-supplied fuel.
Flight 10’s data haul from engine restart dynamics to heat shield stress mapping feeds directly into these ambitions. As acting NASA Administrator Sean Duffy explained, the mission “paves the way for the Starship Human Landing System that will bring American astronauts back to the Moon on Artemis III.” For SpaceX, the issue is no longer whether or not Starship will survive a full test profile, but how rapidly it can iterate towards operational capability. Flight 11 will be the next gauge of that rate.
