The FAA Says SpaceX Can Fly Starship Again
The Federal Aviation Administration handed SpaceX the green light on Monday, July 14, 2026 — clearing the company to resume Starship test flights after identifying the probable cause of its Super Heavy booster's catastrophic failure during a May 2026 flight. SpaceX announced over the weekend that the next launch could happen as soon as Thursday, July 16. It would be the second-ever flight of Starship Version 3 (V3), and this time it's carrying real third-generation Starlink satellites instead of the dummy payloads used previously.
This is SpaceX's second test flight of Starship and its first as a public company. The IPO closed on June 12, raising nearly $86 billion on the Nasdaq and making SpaceX one of the ten most valuable companies in the world. The market clearly bought into the "fly, fail, fix" philosophy that CEO Elon Musk has championed — though he prefers the term "rapid unscheduled disassembly" for when things go wrong.
The FAA's clearance process here mirrors the kind of rigorous oversight NASA applies through its Commercial Crew Program. When NASA certifies SpaceX's Crew Dragon or Boeing's Starliner for human spaceflight, it demands the same kind of root-cause analysis and hardware verification that the FAA is now requiring for Starship's booster. The difference is scale: this isn't about carrying humans yet, but the engineering rigor is no less demanding. SpaceX had to prove it understood exactly why its booster failed, fix the underlying issues, and demonstrate those fixes would hold under flight conditions. The FAA was satisfied.
What Happened During the May 22 Flight
SpaceX's first V3 test launch on May 22, 2026 was largely successful — at least until it wasn't. The Super Heavy booster lifted the 407-foot rocket into space, the upper stage separated cleanly, and SpaceX deployed twenty satellite simulators along with two modified Starlinks that recorded footage of the Starship exterior. The third-generation booster was supposed to return to Earth and perform a simulated landing in the Gulf of Mexico.
Instead, its engines didn't properly re-ignite. The booster plummeted into the water below.
According to both SpaceX and the FAA, the problem occurred at booster separation. "Slight differences in engine startup on the ship" caused the Booster to turn ninety degrees in the wrong direction. That misalignment cascaded into a failure where the booster couldn't relight its engines for the simulated Gulf landing. The FAA's investigation identified two most probable root causes: "heat effects on propulsion system components during ascent" and "erroneous engine alarm system settings." In other words, the extreme thermal environment during the rocket's climb through the atmosphere affected components in ways SpaceX hadn't fully anticipated, and the alarm system was set up in a way that either masked or misreported the developing problem.
The upper stage itself had its own issue. While it successfully deployed its test payload and simulated a Gulf landing, it lost one of three Raptor engines meant for vacuum use. SpaceX said it made "several hardware and operational modifications" to prevent that engine loss from recurring.
The Fixes: What SpaceX Changed Before Flight 13
SpaceX didn't just file a report and wait for permission. It went to work.
The company modified the engine startup sequence so the booster can "more reliably flip in the desired direction" after separation. It also modified the booster itself to "improve re-light reliability" — meaning the engines that need to fire again during the descent and simulated landing phase are now more likely to actually light when commanded. SpaceX changed Starship's engine alarm and abort systems to reduce the chance of a similar failure, addressing the FAA's concern about "erroneous engine alarm system settings."
For the upper stage, SpaceX made "several hardware and operational modifications" to prevent the vacuum Raptor engine loss from recurring. These aren't cosmetic tweaks. We're talking about changes to how the propulsion system handles the thermal environment during ascent, and how the alarm architecture interprets sensor data in real time. That's the kind of deep engineering work that separates a company that learns from failure from one that just repeats it.
The "fast-fail, fast-fix" cycle is what makes SpaceX different from traditional aerospace programs. NASA's Commercial Crew Program took over a decade and billions in cost growth before achieving certification. SpaceX identified the booster failure, diagnosed it, fixed it, and was back in front of the FAA within two months. That pace is why investors piled into the IPO and why competitors are watching nervously.
Flight 13: The July Timeline and the T-0 Abort
Flight 13 is the thirteenth test launch of Starship since 2023, and it almost didn't happen on schedule.
SpaceX first attempted to launch Flight 13 on Thursday, July 16 at 6:45 p.m. EDT from Starbase, Texas. The vehicle was stacked with Booster 20 (Super Heavy) and Ship 40, both of which had completed static test fires of their Raptor 3 engines. The automated flight computer initiated a hold and abort call at T-0 when several of the thirty-three Raptor engines failed to ignite as planned. SpaceX livestream host Dan Huot confirmed: "We did trigger a hold on the booster that aborted our liftoff as we were starting to light those Raptor engines."
Musk posted shortly after: "To be confident of a good flight, 2 Raptors will be removed & replaced." At least three Raptor 3 engines were ultimately swapped out. Booster 20 returned to the pad on July 22 for additional preflight testing, which SpaceX said went well. The new target was a 90-minute launch window opening at 6:45 p.m. EDT on July 23.
The abort was a reminder that even with FAA clearance, the hardware can still throw curveballs. But it also demonstrated the system working as designed: the automated flight computer caught the anomaly, called the abort, and SpaceX responded methodically rather than rushing. That discipline matters — especially when you're flying a $1.8 trillion company's most important vehicle.
The Payload: First Real Starlink V3 Satellites in Space
This flight carries something Flight 12 didn't: twenty functional third-generation Starlink satellites. Previously, Starship had only carried dummy versions of the larger, more powerful internet satellites.
These new satellites are designed to connect with the broader Starlink constellation "via high-capacity lasers" and will burn up in the atmosphere roughly twenty minutes after deployment. Six of them are equipped with cameras specifically tasked with photographing Starship's exterior during the flight. SpaceX painted select heat shield tiles white to create target markers, turning the satellites into a controlled test environment for monitoring how those tiles react to the stresses of exit and reentry.
It might seem counterintuitive to launch expensive tech on a test flight destined for a splashdown, but the logic is sound. These satellites aren't meant to hold orbit. They're high-tech test subjects, gathering critical environmental data that would be impossible to get from ground-based testing alone. Starlink was the only profitable part of SpaceX's business ahead of its IPO, so every piece of hardware that improves network capacity and user speeds matters enormously to the company's bottom line.
The V3 versions of both Starship and Starlink are crucial to SpaceX's future. The company needs Starship to become a fully reusable rocket system to attempt its plans for space-based data centers and interplanetary travel. Flight 13 is a step toward that goal — not the final one, but a necessary one.
Why This Matters Beyond the Launch Pad
The FAA's clearance of SpaceX's Starship program sits at an interesting intersection of regulatory philosophy and commercial reality. NASA's Commercial Crew Program was built on the premise that private companies could develop crew transportation systems more efficiently than government programs — but it still demanded exhaustive certification before allowing humans aboard. The FAA is applying a similar standard here, just for an uncrewed vehicle carrying increasingly valuable payloads.
The May failure and the subsequent fixes demonstrate that this model works. SpaceX failed, the FAA investigated, SpaceX fixed the root causes, and the regulator was satisfied. The two-month turnaround from failure to re-clearance is faster than most traditional aerospace programs achieve in a single development cycle. That's the promise of commercial spaceflight: iterate quickly, fail cheaply, learn faster.
Flight 13 itself — with its T-0 abort and engine swaps — proved the model holds even when things don't go perfectly. The automated systems caught the anomaly, SpaceX responded methodically, and the launch window remained open for another attempt. No panic. No rushing. Just engineering discipline under extreme financial pressure.
For the aerospace industry at large, this is a benchmark. The Commercial Crew Program took over a decade to certify Crew Dragon. Starship's booster failure-to-reclearance cycle was measured in weeks, not years. Whether that pace is sustainable as the program matures remains to be seen, but for now, SpaceX is writing the playbook on how to do rapid iterative development at the scale that matters.