Miami Cargo Jet Crash Kills Five

When a large transport overruns a runway, the human toll usually happens beyond the pavement; the Miami Amazon-branded cargo 767 made that brutally clear, with five people on the ground killed after the jet exited Runway 30 and struck vehicles outside the airfield boundary.

The Short Version

A Boeing 767-300 cargo aircraft operated for Amazon service by 21 Air overran a landing at Miami International, a major accident that killed five people and injured five more.
The aircraft arrived from San Juan and departed the runway environment near the northwest end of the airport before colliding with ground vehicles; the pilot and copilot survived and were later released from the hospital.
The NTSB recovered the flight data and cockpit voice recorders; a full causal analysis will integrate these with performance data, airport conditions, and crew actions.
Runway overruns are a recurring accident class; mature investigations often find a chain of operational, mechanical, and environmental factors rather than a single cause.

What happened on the ground: an overrun that turned into a roadway disaster

The airplane, a Boeing 767-300 freighter operated by Miami-based 21 Air on Amazon’s Prime Air network, arrived from San Juan and landed at Miami International in the early afternoon. After touchdown, the aircraft failed to stop within the available distance, ran off the end of Runway 30, breached airport confines, and struck multiple vehicles. Five people were killed—authorities later specified the victims were on the ground, including occupants of a cleaning-company van; five additional people were injured, three critically. The pilots survived, received treatment, and were discharged in the following days. The Federal Aviation Administration characterized the event as an overrun occurring around 2 p.m. Eastern time and the Miami-Dade Aviation Department reported the jet disabled at the northwest end of the airfield.

The human pattern here is tragically familiar: the kinetic energy an airliner carries at landing is enormous; if it does not stay on the runway, the consequences are borne by whoever or whatever lies beyond the threshold. In Miami, that meant road traffic and support vehicles in the airport’s perimeter—people with no role in the flight who suffered the worst of the outcomes.

How investigators will build the causal chain

Within 24 hours, National Transportation Safety Board specialists recovered both the cockpit voice recorder (CVR) and flight data recorder (FDR), the paired instruments that anchor modern accident reconstruction. The FDR will show touchdown point, groundspeed, brake and spoiler deployment, reverse-thrust commands, autobrake settings, and deceleration rates; the CVR aligns crew callouts, automation mode awareness, and workload management to those data. Together they will establish whether the stopping problem began before touchdown—excess approach speed, float, or a late touchdown—or after touchdown due to configuration, system issues, runway condition, or reverse-thrust timing.

Expect the NTSB to layer in additional evidence: airport surveillance video and ground-scar mapping, meteorological data for wind and surface condition at landing, runway friction testing, maintenance and minimum equipment list (MEL) records, and crew qualification, duty, and rest histories. The result is rarely a single-cause story; it is an interaction of human and machine, of performance margins and decisions made in seconds.

Why runway overruns keep happening: mechanism and precedent

Runway excursions are a stubborn, well-studied risk category because the physics is unforgiving. Stopping distance is set by kinetic energy (which scales with the square of speed), the friction the tires can generate after lift is dumped, the contribution of reverse thrust, and runway length remaining at touchdown. Three mechanisms dominate accident reports: arriving too fast or floating and touching down long, failing to get spoilers and reverse thrust in promptly to transfer weight to wheels and maximize braking, or degraded stopping performance from contamination or mechanical deficits. The nuance is that several can be true at once.

NTSB precedent underscores that variety. The Board has traced overruns to pilot technique and approach discipline, as in the Embraer 170 overrun at Cleveland where a contaminated runway met an unstabilized approach and delayed braking actions, and to hidden mechanical contributors, such as the PenAir SA-2000 whose incorrectly wired anti-skid sensors compromised braking and led to a fatal overrun at Unalaska. Maintenance-driven tire underinflation has also precipitated catastrophic loss of stopping capability, as documented in a prior MD-80 case. These cases matter because they prevent premature compression of the Miami crash into a single-culprit narrative before the data is read.

Amazon cargo jet tried to abort unsteady landing before Miami crash that killed 5, per flight data https://t.co/a5K63Svk8o

— WMBF News (@wmbfnews) September 9, 2026

The Miami specifics: aircraft, operator, and runway environment

The involved airframe was a Boeing 767-300 freighter on Amazon-branded service; branding can obscure accountability, but for regulatory purposes the operator of record—21 Air—controls crew training, dispatch, maintenance, and operating procedures. The airplane reportedly overran after landing on Miami’s northwest alignment; public flight-tracking snapshots reported high groundspeed near runway end, but those preliminary figures will be replaced in the docket by calibrated FDR values with precise timestamps. Former officials have noted that most U.S. runways do not have engineered materials arresting systems (EMAS) at every end; runway safety relies instead on disciplined landing performance planning, stabilized-approach criteria, and immediate deployment of lift-dump and deceleration devices on touchdown.

What to watch as facts harden

Several technical questions will determine the causal spine. First, touchdown point and speed: did the airplane cross the threshold fast or float past the aiming point such that remaining length could not absorb the required energy? Second, lift dump and deceleration: were ground spoilers armed and did they deploy immediately with weight-on-wheels on both main gear, and when did reverse thrust reach effective levels? Third, runway state: was the surface dry, damp, or contaminated, and did any tailwind, gust, or shear reduce stopping performance? Fourth, systems and maintenance: did anti-skid, brake, or thrust-reverser anomalies reduce effectiveness, and were there any MEL deferrals? Finally, crew factors: were approach gates respected, was a go-around considered when parameters diverged, and did duty time or task saturation degrade performance? The FDR and CVR carry decisive answers to each of these.

Why this matters beyond one runway

Every high-visibility overrun reverberates through airline training, airport risk management, and community safety. Operators revisit stabilized-approach enforcement and landing-distance assessments; dispatchers and crews recalibrate conservatism on shorter or condition-limited runways; airports reassess end-of-runway hazards, vehicle access controls, and the feasibility of EMAS at threat-heavy runway ends. Regulators and manufacturers use the data to tune checklists, annunciations, and automation modes that minimize delay in spoilers and reversers when seconds determine outcomes. For communities bordering major hubs, the stakes are immediate: what sits beyond the fence—roadways, parking lots, work vans—should be planned with the rare but catastrophic overrun in mind.

Bottom line

The Miami accident’s core facts are firm: a 767 freighter overran a landing, five people on the ground were killed, and the flight crew survived. The recorders are in hand, and the NTSB’s methodical reconstruction will convert speculation into sequence. History says the answer will be a chain, not a headline. That is how complex systems fail—and how aviation learns.