Air Florida Flight 90 — Iced Wings, Anti-Ice Off, Down on the 14th Street Bridge
Summary
On 13 January 1982, Air Florida Flight 90, a Boeing 737-200 attempting to take off from Washington National Airport in a snowstorm, climbed only a few hundred feet before stalling, struck the 14th Street Bridge over the Potomac River, and fell into the freezing, ice-choked water. Of the 79 people aboard, 74 passengers and 5 crew, all but five were killed; four people on the bridge also died in the impact. The official toll was 78 dead — 74 aboard the aircraft and 4 on the ground — with five survivors pulled from the river. It was a takeoff accident in plain sight of the United States capital, and the National Transportation Safety Board's reconstruction made it one of the most-studied crew-performance cases in aviation.
The 737, registration N62AF, had been deiced before pushback, but a long delay between deicing and departure left it accumulating fresh snow and ice on the wings as it waited in the falling snow for takeoff clearance. Critically, the crew did not switch on the engine anti-ice system. Without it, the engine pressure probes iced over and gave falsely high thrust readings; the engines were in fact producing substantially less power than the gauges indicated. On the takeoff roll the captain pressed on despite the first officer twice voicing concern that the readings looked wrong. Contaminated by ice and under-powered, the aircraft lifted off, struggled to climb, stalled, and came down on the bridge and into the river.
The NTSB determined the probable cause to be the flight crew's failure to use engine anti-ice during ground operation and takeoff, their decision to take off with snow and ice on the wings, and the captain's failure to reject the takeoff when the first officer drew attention to the anomalous engine readings. Contributing factors included the prolonged ground delay after deicing, the known tendency of the 737 to pitch up when its leading edges are contaminated, and the crew's limited experience operating jet transports in winter conditions.
The crash drove lasting changes in cold-weather operating procedures, in deicing and holdover practice, and in the training of crews for winter takeoffs. It also entered the public record for the conduct of the rescue, including a passenger who repeatedly passed a helicopter lifeline to others before slipping beneath the ice, and the bystanders and aircrews who pulled survivors from the river.
Timeline
A Capital in the Snow
13 January 1982 was a day of heavy snow in Washington. Washington National Airport, hard against the Potomac, had closed briefly for snow clearance and reopened into continuing snowfall. Air Florida Flight 90 was a scheduled service to Florida operated by a Boeing 737-200, registration N62AF, with 74 passengers and 5 crew aboard. The captain and first officer had relatively little experience of jet operations in genuine winter conditions, a fact the investigation would later weigh; Air Florida was a warm-weather carrier, and snow-and-ice takeoffs were not its routine environment.
The aircraft was deiced at the gate before pushback. But deicing fluid buys only a finite holdover time, and on a heavily snowing day that margin shrinks fast. After deicing, Flight 90 waited — through pushback difficulties and a queue for takeoff — while snow kept falling on its wings. By the time it reached the runway, fresh contamination had built up on the lifting surfaces, the very condition deicing had been meant to remove.
A second, quieter error compounded the first. The crew did not turn on the engine anti-ice system, which keeps the engine inlet pressure probes clear of ice. With those probes iced, the engine instruments would read thrust higher than the engines were actually producing. The gauges would show the crew a takeoff that was going well while the engines were in fact delivering far less power than the numbers implied. The aircraft was being set up to attempt a takeoff that was both aerodynamically contaminated and quietly under-powered, and the cockpit had no clear indication of either.
The Takeoff That Would Not Climb
On the takeoff roll, the cockpit voice recorder captured the first officer's unease. He twice remarked that the engine readings did not look right — that something was wrong with the way the aircraft was accelerating relative to the indicated thrust. The captain, with the takeoff already under way, did not reject it. The 737 lifted off, but it was in trouble from the moment it left the ground. The contaminated leading edges disturbed the airflow over the wings, and the 737's known tendency to pitch up when its leading edge is contaminated worked against the crew. The under-powered engines could not supply the thrust to overcome the added drag and the degraded lift.
The aircraft staggered into the air, reached only a few hundred feet, and could not climb. It entered a stall — the wings, fouled by ice and starved of the necessary airspeed, simply stopped producing enough lift. Settling rather than flying, the 737 came down onto the 14th Street Bridge, which spans the Potomac near the airport. It struck vehicles on the bridge, killing four people on the ground, and then plunged through the ice into the river.
Most of those aboard died in the impact or in the freezing water. Five survived, pulled from the river by helicopter crews, emergency responders, and bystanders who went into the cold themselves. The rescue became part of the historical account: a passenger in the water repeatedly handed the helicopter's lifeline to others rather than taking it himself, and was gone by the time the helicopter returned for him. The conduct of those who helped, and the conditions they worked in, were documented soberly in the record of the day.
The NTSB Reconstruction and Its Probable Cause
The National Transportation Safety Board investigated and adopted its report, AAR-82-08, in August 1982. Its probable-cause statement was specific and direct. The Board determined that the probable cause of the accident was the flight crew's failure to use engine anti-ice during ground operation and takeoff, their decision to take off with snow and ice on the airfoil surfaces of the aircraft, and the captain's failure to reject the takeoff during the early stage when his attention was called to the anomalous engine instrument readings. The finding was a crew-performance finding — pilot, in the Board's attribution — built from three linked decisions, each of which alone might have been survivable.
The Board did not treat the crew in isolation. It cited contributing factors: the prolonged ground delay between deicing and the receipt of takeoff clearance, during which the aircraft sat in continual precipitation; the known inherent pitch-up characteristics of the 737 when its leading edge carries even small amounts of snow or ice; and the limited experience of the flight crew in jet-transport winter operations. The picture the report assembled was of a crew unfamiliar with winter flying, operating an aircraft with a known icing sensitivity, who let the deicing margin lapse, omitted the anti-ice that would have kept their instruments honest, and pressed on through a warning the first officer voiced aloud.
The accident became a foundational case study in crew resource management and in cold-weather operations. The first officer's unheeded concern on the takeoff roll is a textbook illustration of a junior crew member's warning that did not change the captain's decision, and of how authority gradients can suppress the cross-check that should stop an unsafe takeoff.
The Five Factors
Aftermath
Air Florida Flight 90 reshaped winter flying. The accident drove revised cold-weather operating procedures across the industry: stricter discipline on deicing and holdover times, clearer rules on the clean-aircraft requirement before takeoff, mandatory use of engine anti-ice in icing conditions, and improved training for crews operating in snow and ice. It reinforced, alongside other accidents of the era, the crew-resource-management movement — the recognition that a first officer's unheeded warning is itself a failure mode, and that procedures must let a junior crew member stop an unsafe takeoff. The case remains a fixture of pilot training in icing and in cockpit decision-making.
The human record of the day endured as well. The five survivors were pulled from the Potomac in extreme cold by helicopter crews and civilian rescuers, and the passenger who repeatedly relinquished the lifeline to others before he was lost became one of the most-remembered figures of the rescue. The 78 dead — 74 aboard and four on the 14th Street Bridge — and the conduct of those who tried to save them are both part of how the accident is recorded.
Lessons
- Treat deicing holdover time as a hard operational limit; a long ground delay after deicing in continuing snow can rebuild the contamination the treatment removed.
- Use engine anti-ice whenever conditions require it; without it, iced pressure probes can make thrust gauges read high and hide an under-powered takeoff behind good-looking numbers.
- Honour the clean-aircraft principle absolutely — even small amounts of snow or ice on the wings can be disqualifying, especially on types with a known leading-edge sensitivity.
- Empower any crew member to reject an unsafe takeoff; the first officer's twice-voiced concern was the missed signal, and a steep authority gradient is itself a hazard.
- Train crews for the conditions, not just the aircraft; operating in a winter environment outside a carrier's normal climate demands specific cold-weather procedure and practice.
References
- Air Florida Flight 90 — accident report AAR-82-08 (PDF) National Transportation Safety Board (NTSB)
- NTSB accident docket DCA82AA011 (Air Florida Flight 90) National Transportation Safety Board (NTSB)
- Air Florida Flight 90 Wikipedia (synthesis of NTSB report AAR-82-08 and contemporary reporting)