“What does it take to light a rocket engine in the vacuum of space and do it twice?” For SpaceX, the answer to that question could define the next chapter of American lunar exploration. On August 24, the company plans to launch Starship Flight 10 from its Starbase facility in Texas, a mission carrying not just hardware but the weight of NASA’s 2027 Artemis 3 deadline.
At approximately 394 feet tall, fully stacked Super Heavy booster and Starship upper stage will make history on the Block 2 configuration of Booster 16 and Ship 37. These models include a sequence of purposeful upgrades from previous flights, based on lessons gleaned from a series of partial successes and downright failures. Block 2’s engineering upgrades are not superficial. They feature a 25% increase in propellant capacity 3,650 tons for the booster and 1,500 tons for the ship increasing payload capacity to more than 100 metric tons to low-Earth orbit when they are reused. Structural upgrades, including thinner, relocated forward flaps, are intended to enhance aerodynamics and make it easier to construct heat shields, while the use of both Raptor 2.5 and Raptor 3 engines obviates secondary engine shielding.
The Raptor engine itself continues to be at the heart of the mission risk profile. Built for deep-throttle ability and restart in microgravity, the methane-burning powerplant has been uncooperative in vacuum environments. Past NASA astronaut and SpaceX consultant Garret Reisman explained them in blunt terms: They’re finicky little beasts and it’s not so easy to light them up and shut them down and light them up again. Flight 10’s in-space relight test is more than a technical formality; it is required for controlled reentry from orbital missions, including the return from the lunar surface.
Mission objectives go beyond propulsion tests. After stage separation, Ship 37 will release eight Starlink satellite mass simulators to test payload deployment systems, where the door mechanism jammed on Flight 9. The ship will next proceed with a controlled splashdown in the Indian Ocean, testing Block 2’s aerodynamic and thermal protection improvements under actual reentry loads. Booster 16, in the meantime, will aim for a standalone controlled descent into the Gulf of Mexico, skipping a “Mechazilla” tower catch. Chopstick arms of the tower, designed to support a falling 230-foot booster, have proved their accuracy in previous flights, but automated health checks also initiated last-minute aborts, highlighting the intricacy of the system.
The journey to this launch has not been smooth. May’s Flight 9 concluded with the second stage disintegrating 45 minutes into the flight. Later in the weeks, a static fire test destroyed Ship 36, the initial Flight 10 contender, and also damaged ground equipment. Investigations, carried out by SpaceX under FAA supervision, centered on aerodynamic stability, heat shield behavior, and propellant system irregularities. The Block 2 design is formulated to overcome these weaknesses with enhanced heat shield tiles, increased data pathway redundancy, and a change in design philosophy to focus on mass efficiency rather than sheer volume.
The stakes are heightened by NASA’s dependence on Starship as the Human Landing System for Artemis 3. The agency’s timeline is already tight, with Artemis 2 a crewed lunar flyby planned for April 2026. A slip in Starship readiness could cascade into mission timing, which might prompt NASA to look elsewhere, including Blue Origin’s Blue Moon lander. But a change of providers would be expensive and logistically difficult, particularly with China ramping up its own lunar effort toward a 2030 crewed landing.
From the engineering perspective, Flight 10 is as much a matter of operational tempo as it is about milestone accomplishments. SpaceX has approvals to launch up to 25 Starships in 2025, a rate of activity that, if maintained, might iterate hardware and software at unprecedented rates. Aggressive reuse, facilitated by mechanisms such as Mechazilla and booster recovery, is key to pulling the costs of launches down from hundreds of millions toward the company’s target $15–$100 million per flight.
If successful, the mission will represent a convergence of incremental engineering progress: increased thrust from 8,240 metric tons at liftoff of the boosters, more efficient aerodynamics, and proven in-space restart capability of the engines. Each represents a distinct technical accomplishment, but together they constitute the foundation of a vehicle architecture capable not only of the Moon, but Mars and beyond.
