SpaceX Wants to Leave Starship in Space for the First Time

SpaceX will try to leave its Starship upper stage in orbit around Earth for the first time on Sept. 22. The launch window lasts 75 minutes and opens at 7:15 a.m. CT, according to TechCrunch.

This is the 14th test flight of Starship, SpaceX's large rocket system. Past tests used a shorter path that went to space and came back down without circling the planet. This time the plan is to give the upper stage enough speed to keep circling Earth.

The flight will carry 26 V3 Starlink internet satellites into orbit. That is more than was flown on the prior test.

SpaceX will not try to catch the Starship upper stage with the launch tower on this flight. It will also not try to catch the Super Heavy booster, the large bottom stage that lifts the ship off the pad. Both stages will fly without a tower recovery attempt.

The last flight gives context for that payload. SpaceX's 13th Starship test deployed 20 V3 Starlink satellites, according to Reuters. V3 satellites are a more powerful version with greater bandwidth and internet speeds.

SpaceX has said each Starship launch carrying V3 satellites will add more than 20 times the capacity, according to SpaceX.

Earlier work tested delivery and heat protection. Starship deployed its first batch of mock Starlink satellites in space in August 2025, according to Reuters. That same flight tested new heat-shield tiles during its return through the atmosphere.

The broader context here is that staying in orbit is a different job. It requires steady power and heat control, releasing satellites in order, and planning for safe disposal or return. Shorter flights can check liftoff, stage separation and travel through the air. They cannot check that full loop.

In my view, skipping both catches simplifies this test. It eases range safety and flight operations when the path is new. It also trades data on reuse for data on orbit and delivery. For a rocket meant to fly often, reuse matters. For a rocket meant to build an internet network, reaching a useful orbit with a full load matters first.

Worth flagging for networking and ground teams is what the numbers suggest. Moving from mock satellites to 20 working V3 spacecraft and now to a planned 26 points to steady progress on the dispenser and separation system. If each load does add capacity at that scale, planning for ground stations, backhaul links and service plans changes. Limits would come less from launch pace and more from ground equipment and coordination.

The long arc here is why this flight is worth following, even without a tower catch to watch. A successful trip and delivery would shift Starship from a ship that survives flight to a ship that provides service. We have seen that shift with past launchers. The hardware stops being the story. What it carries becomes the story.