How Safe Is Flying? Aviation Safety By The Numbers In 2026

A Southwest 737 MAX drops to 150 feet over Tampa Bay, four miles short of the runway, and an air traffic controller's warning is the only thing standing between a routine Sunday flight and a catastrophe. The video gets 40 million views. The ATC audio gets clipped, subtitled, and shared across every aviation forum on the internet. Comment sections fill up with theories about pilot competence, DEI hiring, and whether anyone should ever fly a MAX again.

Here's what almost nobody mentions in that conversation: the go-around worked. That's not a footnote. That's the entire point of the system.

I'm not here to defend Southwest's recent run of incidents, several of which deserve real scrutiny from the FAA. But if you're a founder or executive trying to figure out whether flying is actually dangerous in 2026, watching a shaky ADS-B replay of one bad approach is close to useless as a data source. You need the base rates. You need to know what the safety systems actually do, why they exist, and where the real risk sits, because it's very rarely where the headlines point.

So let's do the numbers.

The Actual Fatal Accident Rate, No Spin

IATA's 2025 Annual Safety Report puts the global all-accident rate at 1.13 per million sectors flown, and the fatal accident rate at 0.06 per million sectors, roughly one fatal accident for every 16.7 million flights. To put that in perspective: if you flew once a day, every day, you would need to fly for about 45,700 years before the statistical odds caught up with you.

Fatal accident rates per billion passenger-miles are the more relevant metric for frequent flyers, since a single long-haul flight covers vastly more distance than a regional hop, and distance-adjusted comparisons are how safety researchers actually benchmark risk across transport modes. By that measure, commercial aviation runs at roughly 0.01 fatalities per billion passenger-miles, which is lower than rail, dramatically lower than driving (which sits north of 7.0 per billion passenger-miles in the US), and lower than walking. You are, by a wide margin, safer in the air than you are in the back of an Uber on the way to the airport.

None of this is new. Aviation has been getting safer, almost every year, since the jet age began. What's changed is the noise. A near-miss in 2024 gets filmed by dozens of passengers with phones, uploaded within the hour, and analyzed by YouTube pilots with six-figure subscriber counts before the NTSB has even opened a file. The signal-to-noise ratio on aviation safety has never been worse, even as the underlying safety record has never been better.

What A Close Call Actually Tells You (Spoiler: Not Much)

The Tampa Bay incident is instructive precisely because of how it ended, not how it began. Southwest flight WN425 descended to roughly 150 feet above Tampa Bay, about four miles from runway 10, in heavy rain with strong gusts. Whether the crew mistook the elevated Courtney Campbell Causeway for the runway threshold, as one respected aviation YouTuber theorized after running the ADS-B data through Google Earth, or something else went wrong in the cockpit, we may never know the full mechanism until the NTSB report lands, likely 12 to 18 months out.

What we do know is this: an air traffic controller flagged a low altitude alert, the crew responded within seconds, initiated a go-around, and diverted safely to Fort Lauderdale. The system caught the error. That's not luck. That's redundancy working exactly as designed, layered on top of a human mistake that should never have happened in the first place.

This is the pattern in almost every close call that goes viral. A single link in a long chain fails, sometimes badly, and then multiple independent backup systems catch it before it becomes a fatality. Aviation safety since the 1990s has been built almost entirely around this principle: assume individual humans and individual systems will fail, and build enough redundant, independent layers that no single failure can become a disaster.

That's cold comfort if you're the one white-knuckling the armrest during a go-around. But statistically, a well-executed go-around is a safety system succeeding, not a near miss in the ominous sense the headlines imply.

The Systems Doing The Actual Work: TCAS, EGPWS, RAAS

Three acronyms do more to keep commercial and private aviation safe than almost anything else in the modern cockpit, and none of them get mentioned in viral videos because they're mostly invisible when they work.

TCAS (Traffic Collision Avoidance System) monitors nearby transponder-equipped aircraft and issues automated resolution advisories, verbal commands like "climb, climb now" or "descend, descend now", when two aircraft get too close on a converging path. TCAS II, the version mandated on virtually all commercial and business jets since the mid-1990s, coordinates between both aircraft's systems so that if one is told to climb, the other is automatically told to descend. Midair collisions between TCAS-equipped aircraft have become vanishingly rare as a direct result. The 2002 Überlingen collision, where a Bashkirian Airlines Tu-154 and a DHL 757 collided after a controller's instruction contradicted the TCAS advisory, remains the textbook case study in why crews are trained to always follow TCAS over conflicting ATC guidance. That lesson got written into procedure worldwide.

EGPWS (Enhanced Ground Proximity Warning System) is arguably the single most important safety technology introduced in the last 30 years. It uses a global terrain database combined with GPS position to predict, several seconds in advance, whether an aircraft's flight path will intersect terrain, even in cases where a basic radar altimeter wouldn't catch the danger in time. Controlled flight into terrain, once one of the leading causes of fatal commercial aviation accidents, has dropped by more than 90% since EGPWS became standard equipment. If the Southwest crew's altitude in Tampa had continued its trajectory without ATC intervention, EGPWS would almost certainly have triggered a "PULL UP" warning within seconds.

RAAS (Runway Awareness and Advisory System) is the quieter cousin of the two, designed specifically to prevent runway incursions and wrong-runway landings, the exact failure mode theorized in the Tampa incident. RAAS gives verbal callouts on approach and during taxi: "approaching runway one zero," "on runway two seven," and critically, warnings if an aircraft attempts to land on a taxiway or a runway that's too short for the aircraft type. It's not glamorous. It doesn't get discussed in comment sections. But it's directly relevant to a meaningful share of the incidents that do go viral.

All three systems are standard, non-optional equipment on modern commercial airliners and on professionally operated business jets. Newer fleets tend to carry the latest versions, which is one reason fleet age keeps showing up in the safety rankings below. Part of what a premium ticket on a well-run carrier buys you is exactly this, not just seat width and catering.

Safest Airlines: What The JACDEC Rankings Actually Measure

The JACDEC (Jet Airliner Crash Data Evaluation Centre) safety ranking is the closest thing the industry has to an independent, methodologically consistent airline safety scorecard, and it's worth understanding how it's built before you cite it at a dinner party.

JACDEC doesn't just count fatal accidents per airline, since that would unfairly penalize large carriers with more total flights and reward tiny regional operators that happen not to have had an incident yet. Instead, it weights a composite of factors: fatality-free years, fleet age, incident history (not just fatal accidents), audit results from bodies like IOSA (IATA Operational Safety Audit), and the airline's home country aviation oversight quality.

Consistently, the top of the JACDEC rankings has been dominated by the same names for years: Cathay Pacific, Qantas, Emirates, Singapore Airlines, EVA Air, Etihad, and the major Scandinavian and Gulf carriers tend to trade positions near the top. Qantas famously went decades in the jet era without a fatal hull-loss accident, a record that's practically mythologized in aviation circles at this point, and it holds up under scrutiny.

What's more useful than memorizing a top-ten list is understanding what actually predicts safety at the operator level: fleet age (newer aircraft have more redundant systems by default), regulatory environment (US, EU, and a handful of Asian and Gulf regulators maintain the strictest oversight), and a carrier's willingness to report and act on near misses rather than bury them. Airlines with a genuine safety culture over-report minor incidents internally, because catching small problems is how you prevent big ones. Paradoxically, an airline with a high volume of reported minor incidents is sometimes the safer bet than one that reports almost nothing.

The Risk You're Actually Ignoring: Turbulence

Here's the uncomfortable truth buried under all the crash statistics: turbulence, not mechanical failure, not pilot error on approach, not midair collision risk, is the leading cause of passenger injury in commercial aviation, by a wide margin.

The FAA reports that turbulence-related injuries account for the largest share of airline incident reports involving passenger harm each year, and the overwhelming majority of serious turbulence injuries happen to people who weren't wearing their seatbelt when the seatbelt sign was off, or who were up and moving in the cabin during unexpected clear-air turbulence, the kind that doesn't show up on weather radar and can't be predicted more than a few minutes in advance.

The Singapore Airlines flight SQ321 incident in May 2024, which hit severe clear-air turbulence over Myanmar and resulted in one fatality and dozens of injuries, some serious, wasn't an outlier because of the severity of the turbulence. It was an outlier because of how rarely such incidents result in a fatality at all, precisely because modern aircraft are engineered to handle turbulence loads far beyond what passengers experience as "the plane is going to fall apart" levels of shaking. The airframe was never in danger. The people not buckled in were.

This is, statistically, the single most actionable safety takeaway in this entire piece: keep your seatbelt loosely fastened any time you're seated, turbulence or not, sign on or off. It costs you nothing and it is, by a significant margin, the highest-leverage safety behavior available to any passenger, commercial or private.

Private Aviation: A Different Risk Profile, Not Necessarily A Worse One

There's a persistent myth that private jets are statistically more dangerous than commercial airliners. It's not entirely wrong, but it's not applied correctly either, and the nuance matters if you're deciding how to move around this year.

General aviation as a broad category, which lumps together everything from a Cessna 172 flown by a weekend hobbyist to a Part 91 owner-flown turboprop, does carry a materially higher accident rate than scheduled commercial aviation. But that category is not representative of Part 135 charter flying on professionally crewed jets, which is a different regulatory and operational world entirely. Part 135 operators are subject to specific FAA oversight, mandatory crew rest rules, recurrent training requirements, and, for reputable operators, third-party audits through ARGUS or Wyvern that go beyond baseline FAA minimums.

The honest comparison: professionally operated Part 135 jet charter, flown by two-pilot crews on aircraft like a Citation Latitude or a Challenger 350, sits much closer to scheduled airline safety statistics than it does to owner-flown piston general aviation. The gap that remains is mostly about redundancy at scale (airlines run thousands of flights a day feeding a massive statistical base, and different aircraft types and route networks) rather than any fundamental difference in how seriously safety is taken in the cockpit.

If anything, the routing flexibility of private charter cuts against one of aviation's more persistent risk factors: weather-driven diversions and go-arounds at high-traffic hub airports during bad weather, exactly the conditions that preceded the Tampa Bay incident. A charter operator can choose a different arrival airport, a different approach time, or simply wait out a weather cell in a way that a scheduled airline juggling gate slots and connecting passengers often can't. That's not a safety guarantee. It's a flexibility advantage that, used well, reduces exposure to the exact scenario that made the Southwest incident newsworthy in the first place.

What The Statistics Actually Tell You If You Fly Often

If you're on a plane 40, 60, 100 times a year, which describes a meaningful share of the CEOFLIGHTS audience, the math changes slightly from the casual traveler's, but not in the direction most people assume.

More flights means more cumulative exposure to the base rate, yes. But the base rate is so low (again: one fatal accident per roughly 16.7 million flights globally) that even frequent flyers remain, across a flying lifetime, at dramatically lower cumulative risk than the same person's decades of car commutes. Run the numbers: someone flying 80 times a year for 30 years accumulates 2,400 flights. Against a global fatal accident rate of 0.06 per million, that's a lifetime risk so small it doesn't meaningfully register against other everyday risks you don't think twice about.

What frequent flying does increase is exposure to the non-fatal, higher-frequency risk categories: turbulence injury, delay and diversion related stress, and the cumulative fatigue of travel itself, which is a real health variable even if it doesn't show up in NTSB reports. The seatbelt discipline point applies double if you're logging serious annual hours in the air. So does choosing operators and airlines with strong safety cultures over marginal price savings, since the data is clear that safety investment correlates with fleet age, training budgets, and audit rigor, none of which are visible to the passenger in the moment of booking but all of which show up in the long-run accident data.

Reading The Headlines Without Losing The Plot

Go back to the Tampa Bay incident one more time. It's a legitimate story. Southwest has had a genuinely concerning run of incidents over the past two years, and the FAA scrutiny that followed United a few years back is a reasonable comparison point worth raising about Southwest's pilot training pipeline. None of that is wrong to discuss.

But the correct read on that incident isn't "flying is getting more dangerous." It's "a specific operator had a specific failure that a specific redundant system caught, and that failure deserves specific scrutiny." Generalizing from one viral video to "aviation safety is deteriorating" gets the entire statistical picture backwards. 2025 was, by IATA's numbers, one of the safest years in aviation history by several measures. The systems worked. The controller did his job. The crew, whatever went wrong upstream of that moment, executed the go-around correctly when told to.

That's the story that doesn't trend on social media, because a safety system working as designed doesn't make for compelling footage. But it's the more important story, and it's the one the data actually supports.

Bottom Line

Flying in 2026 is safer, by every meaningful statistical measure, than it has been at almost any point in aviation history, and dramatically safer than driving to the airport to catch your flight. The systems built to catch human error, TCAS, EGPWS, RAAS, layered redundancy at every stage, are doing exactly what they were designed to do, even when a single viral video makes it look otherwise.

The real risk sitting in the data isn't a midair collision or a botched approach. It's turbulence, and specifically, turbulence encountered by someone who didn't keep their seatbelt fastened. Fix that one behavior and you've addressed the single largest actionable risk factor available to any passenger, private or commercial.

Choose your airline with the same rigor the data rewards: fleet age, IOSA registration, a regulator with strict oversight, and a genuine safety culture over marginal savings. If you fly private, the equivalent checklist is an ARGUS or Wyvern rating and recurrent training records. The numbers back that choice up, every time. If you want help picking carriers and aircraft with young fleets on the routes you fly most, talk to a CEOFLIGHTS travel manager.