SpaceX intends to attempt a catch of both stages of its Starship rocket on the vehicle's 15th test flight, a step that would mark the first time the company has recovered the upper stage at the launch tower and a precondition for the rapid launch cadence it now says it needs. The plan, disclosed on the company's earnings call, follows the 14th test flight by a little more than three weeks. SpaceX founder Elon Musk said the data gathered from that flight was sufficient to attempt catches of the Super Heavy first stage and the Starship upper stage on the next mission, which the company is targeting for November at its Starbase site in South Texas.

A regulatory hurdle over land

The catch attempt depends as much on paperwork as on engineering. Returning both stages to the launch site requires the vehicle to fly over land, including northern Mexico, which needs regulatory approval. SpaceX president and chief operating officer Gwynne Shotwell said the company has had "significant constructive engagement" with the Federal Aviation Administration and expects the Flight 15 license to include return-to-launch-site and tower catch for both stages. Musk put the odds of success at "more likely than not," while noting the company is hardware rich if either catch fails. Catching the upper stage is the harder of the two: it re-enters from orbital velocity, and the tower arms must capture it without damaging the vehicle.

Reuse is the whole point

Beyond Flight 15, Shotwell said SpaceX will attempt to re-fly a Starship upper stage "early" next year and will seek to complete a long-delayed Starship-to-Starship refueling demonstration early in 2027. That transfer test is a milestone for both SpaceX and NASA, which needs to prove that Starship can be refueled in low-Earth orbit for lunar missions under the Artemis program. Propellant transfer at this scale has never been performed, and the company needs data on transfer efficiency and on the boil-off rate of liquid oxygen and methane in space. Those figures will refine projections for how many tanker launches are needed to support a Starship landing on the Moon. Reuse and refueling are coupled: a lunar mission may require multiple tanker flights, so the economics only work if each vehicle can fly again quickly.

The cadence target

Musk said he expects Starship to reach a weekly flight cadence in the first half of 2027 and to launch twice weekly in the second half of the year. "The rate at which we launch Starship will increase dramatically as we solve rapid and complete reusability," he said, describing a progression in which early flights require some refurbishment between launches, with that work eventually eliminated so a rocket can land and launch again "just like an airplane." No orbital rocket has achieved that level of reusability. SpaceX's Falcon 9 recovers its first stage but still needs a second stage built for every flight, and its boosters require inspection and refurbishment between missions. Starship's design, which recovers both stages, is intended to change that calculus.

A company reshaped around compute

SpaceX's near-term Starship manifest is dominated by its own payloads. Shotwell said the company will continue launching Starlink V3 satellites, the second generation of direct-to-cell "mobile" satellites, as well as prototypes for its Starmind data center constellation in 2027. The earnings call made clear how far the company's center of gravity has shifted toward artificial intelligence: officials spent much of the session discussing plans to deploy up to 10 GW of compute on the ground in 2027, and the long-standing talk of settling Mars was largely absent. Yet the orbital data ambitions still rest on Starship flying often enough to put hardware in orbit. Analysts on the call did not raise the controversy over throttling of Starlink capabilities for Ukraine, and neither Musk nor Shotwell addressed it.

What to watch

The November flight will be the clearest test yet of whether SpaceX can close the loop on both stages. A successful double catch would give engineers real data on heat shield and engine condition after re-entry, the inputs that determine how much refurbishment is needed between flights. The refueling demonstration in early 2027 will then determine whether Starship can support lunar missions on the schedule NASA needs. Musk's cadence forecasts have historically run ahead of reality, so the weekly-launch target is best read as a statement of intent rather than a schedule. But the engineering direction is unambiguous: every milestone SpaceX now describes depends on turning Starship from an experimental vehicle into one that flies, lands and flies again.