Four Raptor 3 engines on Starship's Super Heavy booster refused to ignite at T-0 on July 16, 2026, and the vehicle's flight computer shut the whole stack down before liftoff. Of the 33 engines that were supposed to fire, only 29 did. Engines E12, E13, E5, and E6, telemetry pinpointed them, simply did not start. The countdown had run clean all day. The tanks were fully loaded with 11.5 million pounds of liquid methane and liquid oxygen. Then, at T-0, the automated abort triggered. Starship Flight 13, the third flight of the V3 booster and upper stage, never left the pad. This is the first time in the entire program's history that a full-stack ignition abort has occurred.
The immediate consequence is simple: Starlink V3 satellites stay on the ground. The upper stage's primary payload was 20 V3 satellites bound for the Starlink constellation. V3 is not an incremental upgrade. These satellites are physically too large to fit inside Falcon 9's payload fairing, Starship is the only vehicle in SpaceX's fleet that can deploy them economically. That constraint matters because Starlink connectivity generated approximately $3.26 billion in revenue in Q1 2026 alone, roughly 69% of SpaceX's total quarterly revenue. The company's margin story, its cash generation, its ability to fund the next round of development, all of it leans on V3 deployment. A week-long slip becomes expensive very quickly.
Elon Musk indicated that two of the flagged Raptors will be swapped out before the next attempt, targeting early next week for a relaunch. The decision to replace rather than investigate-in-place signals confidence that this was a discrete hardware issue, a sensor failure, a valve malfunction, a fuel-line anomaly, not a systemic design problem across the engine fleet. But until those engines have flown, the Raptor 3 architecture carries a new failure mode in the public record. The ignition sequence is supposed to be redundant. Losing four engines at once, across the full booster, during the most critical moment of the flight, the moment when abort is still possible, is not a scenario SpaceX appears to have treated as probable enough to require investigation. It is now.
Behind the headline delay sits a second-order consequence that matters more to national space policy than to SpaceX's quarterly earnings. Flight 13's upper stage was supposed to attempt an in-space Raptor engine relight in orbit, a critical engineering test that NASA mandates as a structural prerequisite before Artemis IV can proceed with its lunar lander's propellant-transfer sequence. Starship will eventually refuel itself in orbit before heading to lunar distances. NASA needs to see that capability demonstrated and proven. The relight test was on Flight 13's manifest. It is now delayed. Every technical milestone NASA requires before Artemis IV can fly is now downstream of this engine swap and relaunch.
The regulatory treatment of the abort remains unclear, though the early read is favorable. Because the vehicle never lifted off, the abort triggered at T-0, the stack was undamaged, no flight occurred, it almost certainly does not cross the FAA mishap investigation threshold. The FAA closed its investigation into Flight 12 on July 13, clearing the path for Flight 13. A T-0 pre-flight abort, even one involving four simultaneous engine non-starts, sits in a regulatory gray zone between "incident" and "mishap." SpaceX has not signaled that the FAA will open a new investigation. Assume it does not, unless you hear otherwise.
What to watch: the relaunch date and whether the replacement Raptors complete static fire validation before Flight 13 reattempts. If SpaceX launches early next week, the engine swap proved conclusive. If the schedule slips beyond that, either the root cause was more complex than expected or the validation process surfaced a wider Raptor 3 batch concern. Second, the in-space relight test. Once Starship reaches orbit post-relaunch, the Raptor engine restart in the upper stage becomes the technical proof point that NASA and the industry will scrutinize. That test will determine whether Artemis IV's timeline compresses further.