You watched the last few Starship flights end in fireballs and company shrugs, and you probably figured the "fly it again tomorrow" promise was still a decade out. Flight 13 moved that line. On Friday, July 24, SpaceX lifted its 13th full-scale Starship off Starbase, Texas at 5:51 pm CDT, and for the first time the ship came down in one piece. It tipped over and floated in the Indian Ocean west of Australia instead of burning up the way every Starship splashdown before it had. That one detail is why Elon Musk now says the next flight may try to catch the upper stage with the launch tower.
I have been around enough hardware to know "looks fine" and "is fine" are not the same call, so here is exactly what Flight 13 proved and what it did not.
What actually went right
The ship rode a 408-foot rocket — the V3 version, only its second flight — most of the way around the world and back. The heat shield is the part engineers lose sleep over, because Starship is meant to fly again and again without swapping tiles after every trip. This time, more than 18,000 ceramic tiles held through temperatures near 2,600°F as the vehicle came back through the atmosphere. SpaceX flew drones over the floating ship to inspect those tiles, and by the accounts we have, the shield looked intact. Dan Huot, a SpaceX communications manager on the webcast, called it a "dream scenario" for the team that needs heat-shield data. When your own comms person reaches for the word dream, you know the earlier flights stung.
The cargo was the other quiet win. This was the first Starship flight to carry SpaceX's next-generation V3 Starlink satellites — 20 of them. The company says the new design can push about 1 Tbps of downlink, roughly ten times the V2 satellites, with 160 Gbps of uplink, about 22 times more. Hold those numbers loosely: the satellites rode a suborbital path for the test and burned up about 20 minutes after deploy, so the throughput is a target, not a delivered service yet. One smart touch — SpaceX modified two of the satellites with cameras to scan Starship's heat shield and beam the images to ground. That is the kind of instrumentation that turns a test flight into a data haul.
The booster is still the weak link
Here is the part nobody should skip. The Super Heavy booster under the ship is the bigger, heavier piece, and it is the one SpaceX has caught with the tower arms before. On Flight 13 it finished the high-thrust part of its boostback burn with all 33 Raptor engines for the first time on a V3 booster — real progress — but the engines meant to relight for the landing burn did not all fire. The booster hit the ocean instead of returning to the pad. Same sore spot as the May flight, so it is not fixed.
That gap matters because the tower catch everyone is buzzing about applies to the ship, not the booster, on the next attempt. SpaceX has already caught Super Heavy twice with the Version 2 design. The upper stage comes back slower in some respects but from a far longer trip, and catching it is the true rehearsal for same-day turnaround.
Why the catch is the whole point
Reuse only saves money if you stop throwing the hardware in the sea. A ship that floats in salt water is finished for flight even if it looks whole — saltwater and rocket engines do not mix, and the vehicle was never built to survive a saltwater bath. So the only road to "fly it again tomorrow" is back to the launch mount, not the ocean. Musk's words on X Friday were plain: "Unless we discover problems after mission data review, SpaceX will attempt to catch the ship with the tower on next flight." The mechanical arms on the tower would grab the ship as it slows to a hover, the same trick already used on the booster.
I am not going to call this a sure thing. Mission data review is where the problems hide, and the booster still cannot stick its own landing. But the direction is not subtle: SpaceX is one clean review away from attempting the catch that makes the reuse story real instead of a slogan.
What I would watch before Flight 14
Two checks decide whether the next flight is a milestone or another near miss. First, the heat-shield data review has to come back clean — no cracked tiles, no soft spots under the surface. Second, the booster has to finally relight and land, or the program carries the same gap it had in May. If both hold, catching the ship becomes a live test rather than a slide in a deck.
If you want the deeper background, our breakdown of how Starship works covers the stages and the reuse plan, and our piece on the prior test flight that edged toward Mars missions shows how much the cadence has improved. For the bandwidth side, the Starlink laser backhaul network explains how these satellites talk to each other once they actually reach orbit.
The bottom line: Flight 13 was the first time a Starship came home in one piece, and that is the precondition for everything SpaceX has promised about reuse. Catching it with the tower is the next proof. All gravy if they stick it.
Sources: Ars Technica on the tower-catch plan and Engadget on Flight 13 and the V3 Starlink cargo.