SpaceX has moved the next launch attempt of its Starship rocket to Thursday, 23 July, the latest delay for a flight that has now been scrubbed once and rescheduled twice.
The company had been targeting Monday evening from its Starbase site at Boca Chica in Texas, and said it needed more time to finish work on the vehicle’s propulsion system after last week’s aborted countdown.
The company had opened a 90-minute window at 6.45pm Eastern on Monday before standing down, and has not yet published a precise time for the Thursday attempt.
Starbase sits on the Gulf coast near the Mexican border, where SpaceX assembles and flies the vehicle from its own launch pad.
The site is not universally loved, and a group of 80 nearby residents is suing the company over damage they say the launches have done to their homes.
This is Flight 13, the second outing for the towering V3 version of Starship, and it follows an inauspicious debut.
On Flight 12 on 23 May, the upper stage reached space and released 20 mock satellites before splashing down in the Indian Ocean, while the Super Heavy booster was destroyed shortly after stage separation, denying SpaceX the controlled descent it had wanted.
The trouble this time began on the ground. During the first attempt on 16 July, several of the 33 Raptor engines on the Super Heavy booster failed to ignite at T-0, triggering an automatic abort in the final seconds before liftoff.
Elon Musk said two of the engines would be removed and replaced before the rocket flew again.
SpaceX first pushed the flight to Monday, then to Thursday, saying only that it had modified the propulsion system to address the ignition fault. It has not offered a fuller account of why the work needed the extra days, which is in keeping with a test campaign that tends to announce dates first and reasons afterwards.
The mission itself repeats the profile SpaceX flew in May. Starship is due to deploy the first batch of upgraded Starlink V3 satellites on a suborbital arc, demonstrating the dispenser mechanism and the network’s laser links before the spacecraft splashes down in the Indian Ocean.
The demonstration satellites are not built to stay up, and will burn up in the atmosphere soon after release.
The payload is a piece of the wider plan rather than a prop. The V3 Starlink satellites are larger and heavier than the ones SpaceX has been flying on its Falcon 9 rockets, and the company needs Starship working to launch them at scale, which makes each dispenser test on these flights more than a formality.
SpaceX also plans to attempt an in-space relight of a single Raptor engine, a manoeuvre it needs to master for orbital missions, and has fitted several of the satellites with external cameras to film the heat shield during descent.
There will be no attempt to catch the booster with the launch tower on this flight, a step the company has folded into earlier tests.
The stakes are higher than they were a year ago. SpaceX is now a public company after the largest IPO in history in June, and the roughly $100bn wiped from its market value after the 16 July abort played out in its share price rather than only on the launch stream. Test failures that once drew shrugs now register with shareholders.
There is a manifest waiting on all of this. NASA has picked a version of Starship as the lander that is meant to return astronauts to the Moon under its Artemis programme, and the timeline for that mission moves with every scrub and every lost booster.
The Starship programme has already cost more than $15bn, and the company has said it wants to fly the rocket often enough to resemble an airline schedule rather than a series of set-piece events.
Thursday’s window, if it holds, would be the 13th attempt since 2023, and the first real test of whether the V3 vehicle can string two flights together without losing a booster.
