STARBASE, UNITED STATES — SpaceX's Starship reached orbit for the first time on Sept. 28 and deployed 26 Starlink V3 satellites, although an engine failure forced an early return to the Pacific Ocean.
SpaceX has put its Starship rocket into Earth orbit for the first time, turning its fourteenth integrated flight test into a working satellite-delivery mission despite an engine malfunction that forced the spacecraft to come home hours ahead of schedule. The uncrewed rocket launched from Starbase in South Texas on Monday, September 28, 2026, carrying 26 operational Starlink V3 satellites.
The mission accomplished two objectives that earlier flights had not achieved together: a controlled transition from a safety-focused suborbital trajectory to orbit, and the placement of working satellites into Earth's orbital environment. But a premature shutdown of one upper-stage engine altered the test programme. Rather than remain aloft for nearly ten hours as planned, Starship descended into the Pacific after approximately three hours.
The result gives SpaceX a substantial new body of flight data while leaving an important engineering question unresolved: how reliably can the vehicle operate its engines through the different burns required for longer orbital flights, recovery and future lunar missions?
An Engine Failure Nearly Stopped the Orbital Attempt
The rocket lifted off at approximately 7:48 a.m. Central time. Its Super Heavy booster separated from the Starship upper stage and later completed a controlled splashdown in the Gulf of Mexico. The booster was not scheduled to return to the launch tower, and the upper stage was likewise intended to be expended after its ocean descent.
An engine on Starship shut down earlier than expected during the ascent. That changed the calculations for flight controllers because reaching orbit was not supposed to take priority over retaining the ability to leave it safely. SpaceX communications manager Dan Huot initially told viewers of the company's livestream that the team would not commit to orbital insertion.
The decision changed after engineers examined the remaining engines, particularly those required for later manoeuvres. The spacecraft subsequently executed the additional burn needed to enter orbit. That sequence illustrates why redundancy matters: losing one engine did not automatically mean losing the mission, but continuing required a fresh assessment of the systems still available.
SpaceX later said it had curtailed the orbital portion out of caution following the ascent problem. The vehicle was redirected from a planned splashdown west of Chile to an approved area of the northern Pacific. It ultimately made a controlled descent and touched down near Hawaii. Its subsequent loss in the ocean was not a failed recovery attempt; neither stage was designated for reuse on this flight.
Twenty-Six Satellites Turn a Test Flight Into a Delivery Mission
Reaching orbit was only part of the mission. Starship released 26 operational Starlink V3 satellites at an altitude of roughly 269 kilometres. Unlike the inert or developmental payloads flown on earlier tests, these spacecraft were intended to join SpaceX's broadband constellation. Elon Musk said afterward that all 26 had deployed and were operating normally; SpaceX also reported receiving signals from the satellites.
The deployment matters commercially because Starship is being developed to carry satellites that are larger and more capable than the spacecraft routinely launched aboard Falcon 9. SpaceX says each Starlink V3 satellite can add approximately one terabit per second of network capacity. Multiplying the company's stated figure by the 26 satellites gives a nominal potential addition of 26 terabits per second for this batch.
That is an equipment-capacity figure, not a measurement of the speeds immediately available to customers. The satellites still require initial checks and orbital positioning, while the performance experienced by users also depends on network architecture, ground infrastructure and demand. SpaceX's broader claim is that a Starship batch of V3 satellites could add substantially more capacity than a typical Falcon 9 launch of its smaller V2 Mini spacecraft.
Three satellites in the new batch were also equipped with cameras intended to photograph Starship's heat shield. This arrangement was designed to provide an external view of the spacecraft during orbital operations. Whether every planned inspection image was acquired and returned is a separate question from the confirmed satellite deployment and requires review of the mission data.
Why Starship's First Orbit Is Different From Earlier Tests
Starship has reached space on previous flights, but reaching space and remaining in orbit are different achievements. A suborbital trajectory rises above much of Earth's atmosphere before gravity brings the vehicle back down without requiring a separate deorbit manoeuvre. An orbital trajectory adds sufficient sideways velocity for the spacecraft to continue travelling around Earth until its orbit is altered.
SpaceX deliberately used suborbital trajectories on its first thirteen integrated test flights. The approach allowed engineers to test ascent, staging, payload systems, atmospheric re-entry and controlled descent without depending on a successful engine restart to remove the spacecraft from orbit. Those flights produced valuable information, but they did not demonstrate the sequence required for sustained orbital missions.
Flight 14 crossed that threshold. The insertion manoeuvre showed that Starship could make the transition to orbit; releasing the V3 satellites showed it could use that capability for a commercial purpose. The controlled return demonstrated another essential part of the orbital sequence, although the shortened duration means the flight did not complete its intended six-orbit endurance profile.
The original plan called for roughly six circuits of Earth at an altitude of about 275 kilometres, followed by a deorbit burn and a Pacific splashdown west of South America. A mission lasting about three hours still provides orbital and re-entry information, but it cannot substitute for all the thermal, propulsion and systems observations that a nearly ten-hour flight would have produced.
The Raptor Malfunction Leaves Important Questions Open
Starship's architecture depends on engines performing several distinct jobs. Engines propel the vehicle during ascent, support manoeuvres once it reaches space and will eventually be needed to help recover the spacecraft. For a vehicle intended to be reusable, a successful payload deployment is therefore an important milestone rather than a complete demonstration of the operating system.
The early engine shutdown on Flight 14 also exposed a practical distinction between an engine failure and an overall mission failure. The remaining hardware provided enough capability for the flight team to approve orbital insertion, and the satellite-release sequence went ahead. Nevertheless, the malfunction reduced confidence in continuing the full orbital test, prompting the early return.
Engine data will need to establish precisely why the shutdown occurred and what corrective action, if any, is required before subsequent tests. SpaceX must also assess the performance of the booster's engines during its return and the condition of the spacecraft's thermal protection system. Flight 14 tested multiple systems simultaneously; the success of one objective does not establish that every component performed as designed.
NASA's Updated Moon Programme Raises the Stakes
Starship's development has consequences beyond SpaceX's satellite business. NASA is working with the company on a modified Starship lunar lander, part of the agency's Artemis programme. That spacecraft will require capabilities extending well beyond those demonstrated during Monday's comparatively short mission, including demanding propulsion operations, docking and eventual flight in the lunar environment.
NASA's July 2026 programme update places an Artemis III orbital demonstration in 2027, with SpaceX and Blue Origin developing lander test vehicles for rendezvous and docking activities. The agency's stated objective is to conduct subsequent crewed lunar landings in 2028. Those dates are programme targets, not guarantees that an individual Starship flight will translate directly into a particular launch schedule.
A March 2026 review by NASA's Office of Inspector General also identified lander-development challenges affecting Artemis schedules. The relevance of Flight 14 is consequently specific: orbital insertion and operational cargo delivery are necessary steps in maturing the launcher, while its engine anomaly reinforces the need for reliable repeated manoeuvres before crewed operations can be contemplated.
Starship's proposed lunar role additionally calls for technologies not demonstrated by placing Starlink satellites in low Earth orbit. SpaceX still needs to validate the appropriate configuration and operations for its lunar lander, including the wider mission architecture on which NASA will depend. The latest result is evidence of progress in launch and orbital operations, not proof that the entire lunar transport system is ready.
The Next Test Will Build on a Shortened Flight
The day's central result is clear: Starship reached orbit and delivered 26 operational satellites while recovering from an unexpected engine shutdown. The flight also ended with a controlled ocean descent, rather than a return to the launch site. Those accomplishments are measurable, but so is the difference between the planned ten-hour mission and the approximately three hours flown.
SpaceX's next engineering decisions will depend on its examination of engine telemetry, satellite data, heat-shield imagery and the vehicle's behaviour during re-entry. The company has not established a public timetable for resolving every issue raised by Flight 14, and further tests will be needed to determine how consistently the system can repeat its orbital and recovery operations.
For Starlink, the immediate development is the arrival of a first operational V3 batch delivered by Starship. For the wider programme, it is a transition from proving that the enormous spacecraft can fly to demonstrating that it can perform useful work in orbit, even as its designers continue addressing the systems needed for longer and more demanding missions.