Starship SN4 Goes Out With a Bang After Its Sixth Static Fire
SpaceX's SN4 prototype, the most durable Starship test article yet, was destroyed at the South Texas site when a ground fitting failed post-test.
Well, that’s one way to end a test campaign. Today at SpaceX’s Boca Chica, Texas site, the Starship SN4 prototype went up in a fireball minutes after completing its sixth static-fire test. Nobody was hurt, but the vehicle itself is gone.
SN4 had been the standout of the Starship program so far. Earlier prototypes cracked, buckled, or straight-up popped during pressure testing before ever getting near an engine. SN4 was different — it survived five previous static fires of its single Raptor engine and was widely seen as the first test article solid enough to eventually attempt a real hop test. Today’s test, the sixth, appeared to go fine on its own. The trouble came after, when a ground-side “quick disconnect” fitting — the hardware that connects the test stand’s propellant and pressurization lines to the vehicle — apparently failed. Video from the site shows vapor venting near the base of the rocket before the whole thing erupts in flame, with the top of the vehicle visibly launched into the air by the blast.
It’s worth being clear about what this wasn’t: it wasn’t an engine failure, and it wasn’t a repeat of the pressure-test tank failures that plagued SN1 and SN3. This looks like a ground systems problem — plumbing and fittings at the test stand, not the rocket’s own structure or propulsion. That’s a meaningfully different failure mode, and arguably a more fixable one, since it points at ground support equipment rather than the vehicle’s fundamental design.
Still, it stings. SN4 was the prototype everyone had pinned hopes on for a near-term hop — a short, low-altitude flight test to prove out the vehicle’s basic flight worthiness. That timeline is now obviously pushed back, however long it takes SpaceX to sort out what went wrong with the quick-disconnect hardware and get SN5 or whichever article is next into position.
What strikes me most is how unbothered SpaceX’s whole approach seems to be by exactly this kind of outcome. This is the “build fast, test destructively, iterate” philosophy in action — the same approach that’s produced a string of exploded test tanks over the past year. Elon Musk has talked openly about expecting prototypes to blow up during development. It’s a jarring way to build a spacecraft if you’re used to the more buttoned-up, simulate-everything-first culture at NASA or traditional aerospace contractors, but it’s also produced real hardware progress at a pace few others are matching.
The company already has SN5 and SN6 in various stages of assembly at the same site, so there’s no shortage of test articles in the pipeline. The real question is whether the quick-disconnect failure reveals something structural about how these tests are set up on the ground — if so, expect a pause while that gets redesigned. If it’s a one-off hardware fault, don’t be surprised if the next prototype is on the stand within weeks.
Either way, the Starship program’s pattern holds: build it, break it, learn, repeat. Today was just a particularly dramatic reminder of the “break it” part.