Flight Crew Files
Live ATC Get The Briefing
Home Case Files & Stories Black Box Files Flight Log (News) Community Hub Resources About Live ATC Get The Briefing
Flight Log
UAP FILES: Navy pilots describe a “Tic-Tac” object outmaneuvering an F/A-18 HEROIC: The co-pilot who landed a 767 with zero engine power over Canada BIZARRE: The flight attendant who survived a 33,000-foot fall. Read the case HISTORY: Inside the deadliest day in aviation, two 747s, one foggy runway SCARY: The cockpit voice recording no pilot forgets UAP FILES: Navy pilots describe a “Tic-Tac” object outmaneuvering an F/A-18 HEROIC: The co-pilot who landed a 767 with zero engine power over Canada BIZARRE: The flight attendant who survived a 33,000-foot fall. Read the case HISTORY: Inside the deadliest day in aviation, two 747s, one foggy runway SCARY: The cockpit voice recording no pilot forgets
Case File · Verified Against The Official Investigation

SWISSAIR 111: THE FIRE THEY COULDN'T STOP

An hour out of New York, two experienced pilots smelled something faintly wrong in the cockpit. They did everything the way they had been trained to do it: identified the problem, called it in, chose a diversion airport, ran the checklist. Twenty-one minutes after that first smell, everyone aboard was dead. The fire had been burning above their heads the entire time, in a space nothing on the flight deck could see.

September 2, 1998 New York JFK → Geneva Swissair Flight 111 McDonnell Douglas MD-11 · HB-IWF
Last reviewed and updated: August 2026
229Lives Lost
21 minFirst Smell To Impact
5m 37sFinal Minutes, Unrecorded
0Survivors
The Flight The Checklist The Silence The Recovery The Cause Unanswered Questions The Legacy Sources
Part One

The Flight

Every time in this case file is Atlantic Daylight Time, the local time in Nova Scotia where the flight ended. The official investigation and the flight recorders are timestamped in UTC, three hours ahead. Both are given where the difference matters.

Swissair McDonnell Douglas MD-11, registration HB-IWF, photographed at Zurich in July 1998, seven weeks before the accident
Photo: Aero Icarus via Wikimedia Commons, CC BY-SA 2.0. HB-IWF at Zurich, 14 July 1998 — the actual accident aircraft, seven weeks before the crash.

Swissair Flight 111 was the airline's flagship evening run: New York to Geneva, an overnight hop popular with United Nations staff, bankers and diplomats. It was nicknamed the “UN Shuttle.” On the evening of September 2, 1998, it pushed back from Kennedy Airport at 20:18 New York time with 215 passengers and 14 crew aboard.

In command was Captain Urs Zimmermann, 49, a Swissair training captain with more than 10,000 flying hours. In the right seat was First Officer Stefan Löw, 36. Both men were, by every account in the investigation, exactly the crew you would want in this situation: methodical, unhurried, and well-drilled. Nothing either of them did that night has ever been characterised by investigators as an error of skill or of nerve.

The aircraft was a McDonnell Douglas MD-11, registration HB-IWF, delivered new to Swissair in 1991. It climbed out over Long Island, turned northeast up the coast, and settled into cruise at 33,000 feet on a routing that would take it past Nova Scotia and out over the Atlantic.

22:10:38 · 52 Minutes After Takeoff

An Odd Smell

First Officer Löw remarks on an unusual odour in the cockpit. Nineteen seconds later Captain Zimmermann says “look,” indicating something has become visible. It is not yet obviously smoke.

22:12:15

The Cabin Is Clear

A flight attendant confirms the smell is present on the flight deck but not in the cabin behind it. Twenty seconds later the captain asks whether it is the air conditioning. The first officer answers yes. This is the reasonable conclusion, and it is wrong.

22:13:14

Smoke, Unmistakably

A discernible amount of smoke becomes visible in the cockpit for a second time. Whatever this is, it is not clearing on its own.

22:14:15 · The Call

“Pan Pan Pan”

Flight 111 transmits a Pan Pan — the international radio call for urgency, one step below a Mayday — reporting smoke in the cockpit and requesting a diversion. The crew names Boston, roughly 300 miles behind them.

22:15:06 → 22:15:36

Halifax Instead

Moncton air traffic control offers Halifax, only 66 miles away. The captain accepts. The descent begins. From here the aircraft has sixteen minutes left.

22:17:20 → 22:19:27

Nobody Can Find The Charts

The approach plates for Halifax are not readily to hand — this was an aircraft routed for Geneva, not Atlantic Canada. A flight attendant is asked to move the crew bag containing the charts within the captain's reach.

22:19:37 → 22:19:57

Too High, Too Close

The controller advises that Halifax's Runway 06 requires a back-course approach and that they are 30 miles from the threshold — at 21,714 feet, far too high. The first officer says they need more than 30 miles. They are turned north to lose altitude.

22:20:48 · The Decision That Defines The Case

Fuel Dumping

The crew discusses dumping fuel. The MD-11 is far above its maximum landing weight after only an hour of a transatlantic flight. See The Checklist below.

22:23:45 → 22:24:09

The Aircraft Starts Dying

The CABIN BUS switch is selected off, part of the smoke drill: isolating electrical loads one at a time to find the source. Twenty-four seconds later Autopilot 2 disconnects on its own. It is the first of a rapid cascade of system failures.

22:24:42 · Last Words Received

“We Are Declaring Emergency”

Flight 111 upgrades from Pan Pan to a full emergency. This is the last transmission anyone on the ground receives from the aircraft.

22:25:40 → 22:25:41

The Recorders Stop

The flight data recorder stops. One second later, the cockpit voice recorder stops. Both lose their power supply at roughly 10,000 feet. See The Silence.

22:31:18 · St. Margaret's Bay

Impact

Flight 111 strikes the Atlantic about five nautical miles southwest of Peggys Cove, Nova Scotia. All 229 people aboard are killed instantly.

Official reconstruction of the final flight path of Swissair Flight 111, annotated with cockpit events and a table of time, heading, altitude, ground speed and distance to Halifax
Diagram: U.S. Federal Aviation Administration, via Wikimedia Commons, Public Domain. The official reconstruction of the final track, annotated event by event. Times shown on the diagram are UTC — subtract three hours for the Atlantic Daylight times used in this case file.
Part Two

The Checklist

This is the part of the case that changed aviation, and it is the part that is most often told wrong. The crew of Flight 111 did not waste time. They followed a procedure that had been written, approved and trained for exactly this situation. The procedure was the problem.

Smoke of unknown origin is one of the hardest emergencies in aviation, because the first task is not to fight the fire — it is to find it. The Swissair smoke drill, like most airline smoke drills of the era, worked by elimination: isolate electrical busses one at a time and see whether the smoke stops. It was a long checklist. Running it properly took time.

At the same time, the crew were dealing with an aircraft that weighed far too much to land. An MD-11 an hour into a transatlantic crossing is carrying most of a full fuel load. Landing at that weight risks collapsing the gear or overrunning the runway. So the crew did what the numbers demanded: they turned away from Halifax and out over St. Margaret's Bay to dump fuel, and they began the dump.

Every one of those decisions was correct according to the training and procedures in force that night. Together, they cost the aircraft the minutes it did not have. Investigators would later conclude that the fire had almost certainly already been burning when the first odour was noticed, and that by the time the Pan Pan call went out there may have been no survivable outcome left at all.

…and we are declaring emergency now, Swissair one eleven.

— The last transmission received from Flight 111, 22:24 ADT
What They Could See

Two Instruments And A Smell

Nothing on the MD-11's flight deck monitored the space above the cockpit ceiling. There were no smoke detectors in that void, and none were required. The crew's entire picture of the fire consisted of what they could smell, what they could see drifting into the cockpit, and — in the final ninety seconds — the order in which their instruments began to fail.

What Was Actually Happening

A Fire Feeding On The Aircraft

Above and behind them, insulation blankets sheathed in metallized polyethylene terephthalate — a material certified as fire-resistant under the standards of the day — were burning and spreading flame across the ceiling. The fire was consuming the same class of wiring bundles as it went, which is why the failures arrived in a cascade rather than one at a time.

Part Three

The Silence

At 22:25:40 the flight data recorder stopped. At 22:25:41 the cockpit voice recorder stopped. Impact came at 22:31:18. For five minutes and thirty-seven seconds, the aircraft was still flying and nothing at all was being recorded.

33,000 FT ALTITUDE 22:14 PAN PAN 22:25:41 RECORDERS STOP ~10,000 ft 22:31:18 IMPACT 5 MIN 37 SEC, NO DATA DESCENT PROFILE: SCHEMATIC, NOT TO SCALE
Everything investigators know about the last five and a half minutes of Flight 111 was reconstructed from wreckage, not from the black boxes.

That gap is why this investigation took four and a half years. With no recorder data covering the end of the flight, the Transportation Safety Board of Canada had to rebuild the final minutes physically — from the aircraft itself. More than two million pieces of wreckage were recovered from the seabed of St. Margaret's Bay, and a partial reconstruction of the forward fuselage was assembled in a hangar at CFB Shearwater. Investigators mapped the fire by reading heat damage across thousands of individual fragments: which wires had arced, which materials had melted, which direction the soot had travelled.

It became the largest and most expensive transportation investigation in Canadian history. It is also the reason the case file is as detailed as it is: almost everything in the sections below was established by physical evidence, because the recordings simply were not there.

Part Four

The Recovery

Peggys Cove is a fishing village of a few dozen houses. On the night of September 2, 1998, its boats were the first vessels on scene.

229Total Killed
215Passengers
14Crew Members
2M+Wreckage Pieces Recovered

Local fishermen took their boats out into the dark within minutes of the impact, before any official rescue effort could be organised. They found no survivors. What they found instead stayed with that community permanently, and the strain on the villagers and on the Canadian Coast Guard, navy and RCMP personnel who worked the recovery became its own long aftermath.

Recovery operations continued for more than a year. The seabed was trawled and dredged; the wreckage was barged ashore, catalogued and reassembled. Identification of the victims took months of forensic work, and some remains were never individually identified. Those are interred at Bayswater, on the far side of the bay from Peggys Cove.

Nova Scotia built two memorials, one on each side of the crash site. The Whalesback stands on the granite headland about a kilometre north of Peggys Cove, looking out at the water where the aircraft came down.

The Swissair Flight 111 memorial at The Whalesback near Peggys Cove, Nova Scotia: two curved granite monuments facing each other across a stone plaza
Photo: Doug Kerr via Wikimedia Commons, CC BY-SA 2.0. The Whalesback memorial, north of Peggys Cove. The second stone honours the recovery workers and the community that took the families in.

In memory of the 229 men, women and children aboard Swissair Flight 111 who perished off these shores September 2, 1998. They have been joined to the sea and the sky. May they rest in peace.

— Inscription, The Whalesback memorial, Peggys Cove, Nova Scotia
Part Five

The Cause

The Transportation Safety Board of Canada released report A98H0003 on March 27, 2003, four and a half years after the crash. Its central finding is not about a broken part. It is about what the aircraft was made of.

The Primary Finding

The Insulation Was Flammable

The TSB found that certification standards for material flammability were inadequate: they permitted materials that could be ignited and could sustain and propagate a fire. The thermal acoustic insulation blankets covering the ceiling void were sheathed in metallized polyethylene terephthalate (MPET). Once something ignited them, they carried the fire across the ceiling. This finding, not the ignition source, is the heart of the report.

The Ignition Source

Arcing, Probably In The Entertainment Wiring

Investigators found evidence of electrical arcing in wiring associated with the in-flight entertainment network. But the TSB was careful: it could not determine whether that arc was the “lead event” that started the fire. This is the single most misreported detail of the case — see the note at the end of this file.

How It Got There

An Entertainment System, Bolted On

The IFEN was a seat-back gaming and shopping system — movies, gambling, duty-free — installed under a Supplemental Type Certificate rather than as part of the original aircraft design. It ran on Windows NT, added well over a tonne of weight, and drew power from a cockpit-area bus. The TSB found the approval process for that installation had not adequately assessed how it integrated with the aircraft's existing systems.

Why The Drill Failed

A Checklist Built For A Slower Fire

The smoke procedures in use assumed a problem that could be diagnosed by elimination while the flight continued. They did not assume a fire already spreading out of sight. Nothing in the drill told a crew that the correct response to smoke of unknown origin is to point the aircraft at the nearest runway immediately and sort everything else out on the way down.

Aircraft certification standards for material flammability were inadequate in that they allowed the use of materials that could be ignited and sustain or propagate fire.

— Transportation Safety Board of Canada, Report A98H0003, Finding as to Cause
Case Notes: Unresolved

The Questions That Remain

The material science is settled and the fixes are in place. What people still argue about is the twenty-one minutes.

Could they have made Halifax if they had gone straight in?

This is the question the case is famous for, and the honest answer is that nobody knows. The TSB's own analysis concluded that by the time the crew were aware of a serious problem, the fire had progressed far enough that a successful landing was unlikely. What is certain is that the diversion, the fuel dump and the elimination checklist together consumed the time, and that the procedures of the era actively directed a crew toward all three.

Why did they choose Pan Pan instead of Mayday?

Because at 22:14 that was the accurate call. Pan Pan signals urgency; Mayday signals immediate danger to life. The crew had an odour and intermittent smoke, no fire warning, and every instrument working normally. They upgraded to a full emergency the moment the aircraft began failing around them. The radio call was not the problem — the information available to make it was.

Was it definitely the entertainment system?

Not definitively, and the TSB never said so. Arcing was found in IFEN-associated wiring, and that system was an obvious candidate: it was an aftermarket installation drawing significant power near the area where the fire started. But the report stops short of naming it as the lead event, and honest retellings of this case have to stop there too.

What happened in the last five and a half minutes?

Unknown, and permanently so. The recorders were dead. Wreckage analysis established that the fire continued to spread and that the crew progressively lost flight instruments and control systems. Beyond that, the final minutes of Flight 111 are the largest deliberate blank in this archive.

Part Seven

The Legacy

Swissair 111 is one of the small number of accidents that visibly changed what aircraft are built from. If you have flown on a widebody in the last twenty years, you have flown behind the fixes.

Materials

Mylar Insulation Came Out Of The Fleet

Regulators worldwide ordered the identification and replacement of metallized polyethylene terephthalate insulation blankets across thousands of aircraft, one of the largest retrofit programmes in commercial aviation. MPET blankets are no longer installed.

Certification

The Flammability Test Was Rewritten

The old Bunsen-burner test that had certified those blankets was replaced with a far more demanding radiant panel test that better reflects how a real in-flight fire behaves. Material that passed the old standard routinely fails the new one.

Procedure

Smoke Means Land, Now

Airline smoke and fire checklists were rewritten across the industry to put diversion first rather than last. The modern instruction for smoke of unknown origin is to get the aircraft on the ground at the nearest suitable airport immediately, and troubleshoot on the way.

Oversight

Bolt-On Systems Got Harder To Approve

The Supplemental Type Certificate process — the route by which the entertainment system reached this aircraft — was tightened, with far more scrutiny of how an added system interacts with the electrical architecture it is spliced into.

The Whalesback

Peggys Cove, Nova Scotia

They have been joined to the sea and the sky.

Bayswater

Aspotogan Peninsula, Nova Scotia

Where the unidentified were laid to rest.

Verification

Sources

Every fact in this case file was cross-referenced against at least two independent sources, anchored by the official government investigation. All links open in a new tab.

Official Investigation Report TSB Aviation Investigation Report A98H0003

The Transportation Safety Board of Canada's complete final report on Swissair Flight 111, including all findings as to causes and contributing factors, and the full safety recommendation record.

U.S. Federal Aviation Administration FAA Lessons Learned: McDonnell Douglas MD-11, HB-IWF

The FAA's official case summary and safety-lessons page, including the flight path reconstruction reproduced in this case file.

Public Broadcaster Archive CBC News: Swissair 111, Twenty Years On

Canada's public broadcaster on the recovery, the community of Peggys Cove and the long aftermath for the people who worked the site.

A note on the cause: This case is very widely reported as “a fire caused by the in-flight entertainment system.” That is a compression of what the TSB actually found. Investigators did find electrical arcing in wiring associated with the in-flight entertainment network, and the system's installation was criticised in the report — but the TSB explicitly could not establish that this arc was the event that started the fire. The finding the report leads with is that the aircraft's insulation material was flammable and propagated a fire that should have stayed small. We have followed the report rather than the shorthand.
A note on the altitude: Some retellings give a precise figure for the aircraft's altitude when the black boxes stopped. The TSB records the recorders losing power at approximately 10,000 feet, with the transponder ceasing 23 seconds later at 9,700 feet. We have used the report's own approximate figure rather than a false precision the source data doesn't support.

About The Author

Flight Crew Files Editorial Team

This case file was researched and written by the Flight Crew Files editorial team, led by a former cabin crew member with firsthand experience in commercial aviation. All facts are verified against official accident investigation reports and primary sources.

Learn more about our editorial standards →

You Might Also Like

Black Box Files

More historic cockpit voice recordings and ATC audio: tragedies and miracles, every case sourced and verified.

Explore

Alaska Airlines 261

A worn jackscrew stripped its threads at 31,000 feet, and two pilots fought an unflyable aircraft all the way down.

Explore

Scary Aviation Stories

Terrifying cockpit emergencies, decompressions and close calls pilots never forget.

Explore

JOIN THE DISCUSSION