Resized and converted to WebP. The complete source image is shown. Derivatives use CC BY-SA 2.0.
Geographic context
Accident site near the runway 22L approach lights
Destination: John F. Kennedy International Airport
The accident position is given as 40 degrees 39 minutes north, 73 degrees 45 minutes west. The airport marker gives approximate geographic orientation. Basemap: Natural Earth, public domain.Read supporting evidence
During an instrument approach to Kennedy runway 22L, Eastern 66 encountered a downdraft and a rapid loss of headwind in a strong thunderstorm. Its descent rate increased and its airspeed fell. The aircraft struck the approach light towers, broke apart and burned. Of the 124 people aboard, 113 eventually died.
Read the full incident Follow what happened, from the flight to the investigation findings.
What investigators concluded
The NTSB found that adverse thunderstorm winds caused the high descent rate into the nonfrangible approach light towers. It linked the crew's delayed recognition and correction probably to reliance on visual cues. The adverse winds might have defeated a successful approach and landing even with rapid instrument based responses. Continued use of runway 22L despite the severe weather hazard contributed.
Eastern 66 and the thunderstorm approach to Kennedy
1,806 words / about 9 minutes to read. Based on the investigation records cited below.
From New Orleans to a deteriorating arrival
Eastern Air Lines Flight 66 left New Orleans at about 13:19 eastern daylight time on 24 June 1975. The scheduled passenger service to New York carried 116 passengers and eight crewmembers in Boeing 727-225 N8845E. Four members of the crew were on the flight deck, including a flight check engineer conducting an annual line check of the flight engineer. The crewmembers were qualified, and the aircraft had been maintained in accordance with the applicable requirements.
The flight reached the New York terminal area without reported difficulty. Controllers began vectoring it into the arrival sequence for Kennedy's runway 22L. A broadcast received by the crew described five miles of visibility in haze and an expected instrument landing system approach. Conditions then changed quickly. At 15:52:43, the arrival controller announced that visibility had fallen to two miles in light rain showers and haze, with instrument flight conditions prevailing.
Thunderstorms were affecting the airport's northern side and the approach course. The crew discussed the difficulties of carrying minimum fuel when terminal delays developed and considered the weather at their alternate, LaGuardia. A crewmember referred to about another hour of flying. At 15:59:19, the final vector controller broadcast that a severe wind shift had been reported on final approach.
Flying Tiger Line Flight 161, a DC-8, had already made a difficult landing. Its captain encountered changing winds, turbulence and downdrafts, and needed almost maximum thrust at low altitude to control the descent. After landing, he urged the tower to change to a northwest runway because of the windshear near the ground. The local controller replied that the airport instrument indicated wind aligned with the landing runway. The captain repeated his recommendation.
Eastern Flight 902, a Lockheed 1011, then abandoned its approach. Its crew reported being pulled right and downward, losing airspeed and encountering almost no forward visibility. They applied takeoff power and went around. Flight 66's cockpit recorder captured the report, and its crew confirmed to the final vector controller that they had heard it. The controller cleared Flight 66 for the instrument approach, and the crew said they would report the conditions.
Two aircraft between Flight 902 and Flight 66 also experienced airspeed losses and increased descent rates. Their pilots had anticipated the shear and increased their approach speeds. Both landed and neither sent another windshear report. One pilot believed the controllers already knew about the condition. Flight 66 received Flight 902's warning but never received the Flying Tiger captain's recommendation to change runways.
The first officer was flying. The crew planned an airspeed margin of 10 to 15 knots above the normal approach reference speed. Investigators later considered this margin reasonable given the windshear warning and the risk of excessive landing speed on a wet runway. The captain asked the tower about braking action. The tower reported a wet approach end and said it had received no adverse braking reports.
The aircraft intercepted the glideslope at about 3,000 feet. The crew lowered the landing gear and extended the flaps while completing the checklist. After the first officer requested 30 degrees of flap, the airspeed varied between about 140 and 145 knots as the aircraft followed the glideslope. The flight engineer called 1,000 feet, and rain became audible on the cockpit recording.
At 16:04:38.3, the flight engineer called 500 feet. The aircraft was nearly centered on the glideslope and its airspeed was fluctuating between 140 and 148 knots. Heavy rain followed, and the windshield wipers were set to high speed. The captain directed the first officer to continue using the instruments. He then announced that he could see the approach lights and repeated the instruction to remain on instruments.
As the aircraft passed through about 400 feet, its descent rate increased from an average of approximately 675 feet per minute to 1,500. It moved rapidly below the glideslope. Seconds later, the airspeed fell from 138 to 123 knots in two and a half seconds. At 16:05:06.2, with the aircraft down to 150 feet, the captain announced that the runway was in sight. The first officer responded that he had it. He called for takeoff thrust immediately before impact.
The first impact sound was recorded at 16:05:11.4. The aircraft struck the top of an approach light tower about 2,400 feet from the runway threshold, then hit more towers. The outboard section of the left wing separated. The aircraft rolled sharply left and struck the ground. Its path continued through additional light towers to Rockaway Boulevard, where it stopped. The towers were made of nonfrangible material, and impacts with them and with boulders caused severe destruction of the fuselage.
Fuel escaped after the left wing failed, and fire spread as the aircraft broke apart. Hot engine components, electrical equipment and friction provided multiple possible ignition sources. The two surviving flight attendants had described little or no turbulence during the approach. One recalled a left roll, a large increase in engine power, further rocking, and the emergency lights and oxygen masks before escaping from the wreckage.
The airport fire department responded immediately after the tower raised the alarm. The first firetruck reached the scene about two minutes later. Firefighters brought the main fire under control in about two minutes and extinguished it about three minutes after that. The NTSB credited their rapid intervention with preventing fatal burns among the nine passengers who ultimately survived. Some had been found in pools of fuel and extinguishing foam.
The rear cabin and tail section remained relatively intact, and survivors came from that area. Most passenger seats had torn from their supporting structures as the cabin floors and anchors failed. The aft flight attendants' seats stayed attached, and the attendants escaped unaided. Medical staff gave first aid at the scene. An ambulance took six survivors to hospital, while firefighters transported the others in a firetruck.
Two survivors died shortly after reaching hospital, two passengers died within five days, and another passenger died nine days after the crash. The eventual death toll was 113, leaving eleven survivors. The injury table classified the passenger who died on the ninth day among twelve people with nonfatal injuries because the definition then used covered deaths within seven days. The report's synopsis included the later death in its total.
Both flight recorders were recovered intact and yielded clear recordings. Investigators matched cockpit sounds and radio transmissions with flight recorder traces, radar data and air traffic control time signals. They also examined recordings from aircraft that had preceded Flight 66. Eastern 902's digital recorder supplied more parameters, allowing a more detailed reconstruction of the wind changes encountered along its approach.
The performance analysis indicated that Flight 66 first encountered a growing headwind. As it descended from 600 to 500 feet, the modeled headwind increased from about 10 to 25 knots. Near 500 feet, a downdraft reached about 16 feet per second. By 400 feet, the downward wind had increased to roughly 21 feet per second, while the headwind fell from about 20 to five knots within four seconds. The combined changes reduced lift and drove the aircraft below its intended path.
The wind model fitted the downdraft and outward flowing air associated with strong thunderstorms. Investigators found no malfunction of the aircraft or its components that would have affected its performance, and all three engines were operating normally until impact. The storm's vertical and horizontal air movements were central to the loss of height and airspeed.
Boeing programmed a simulator with the aircraft's flight characteristics and several versions of the reconstructed wind environment. Fourteen pilots flew 54 approaches. Eighteen runs descended to the level of the approach lights, 31 achieved a missed approach, and five reached the threshold in a position to attempt a landing. The tests reproduced rapid deviations below the glideslope near 400 feet and a substantial loss of airspeed.
Pilots often applied less thrust than they believed they had selected. Some hesitated to interrupt their instrument scan to check engine settings, and several used less noseup pitch than the recovery required. The simulator showed that about nine degrees of noseup pitch change could be needed to arrest the descent. The control column force required was greater than several pilots expected after the rapid airspeed loss.
The NTSB concluded that Flight 66's crew probably recognized the developing descent late because heavy rain and poor visibility weakened the visual references on which they were relying. It considered the first officer's transition toward the external view a probable part of that delay. By the time he called for takeoff thrust, the aircraft's high descent rate made impact inevitable.
The NTSB also examined why approaches continued through the thunderstorm. The terminal system was operating at capacity and the controllers were heavily occupied. The tower's local controller said he had no time to pass the Flying Tiger captain's warning to the supervisor responsible for initiating a runway change. He continued to favor runway 22L because the surface wind instrument indicated a wind aligned with it.
At least one northwest runway, 31L, had a relatively clear approach. The Board found that the thunderstorm should have made the danger on the 22L approach apparent. It believed that additional delays and workload discouraged a change. Switching a flight to another runway in instrument conditions required resequencing traffic, and the possible delay also affected the crew's assessment of its fuel reserve.
Weather information had further gaps. A National Weather Service strong wind warning reached airport and airline facilities, but investigators found no evidence that it reached flightcrews. Radar observations showed strong cells merging over the approach course, yet there was no evidence that this information was available to controllers or crews. The NTSB identified continued use of runway 22L in the evident severe weather hazard as a contributing factor.
The NTSB adopted its final report on 12 March 1976. It identified the encounter with adverse thunderstorm winds, producing a high descent rate into the approach light towers, as the probable cause. Its recommendations addressed detection of dangerous winds, the flow of weather information, runway selection and the preparation of pilots and controllers for windshear near the ground.
The Board called for equipment to measure changes in wind along approach and departure paths and for interim surface wind measurements near runway ends. It sought procedures that would give controllers and pilots timely thunderstorm information, runway selection criteria that considered storm location and severity, and training using representative thunderstorm wind models. It also recommended research into airborne alerts and into maintaining vertical guidance during the transition from instruments to an outside view.
The destruction caused by the light towers raised a separate protection issue. Before the accident, the FAA had issued revised criteria requiring frangible structures for future approach light installations. The report recorded that retrofitting existing installations depended on available funds. The NTSB found that the rigid towers struck by Flight 66 had caused extensive damage to the aircraft.
A simplified reading of the investigation findings. The sequence does not replace the full report or imply that every contributing factor is shown.
Sequence of events
Eastern daylight time, using a 24 hour clock. Altitudes are above mean sea level unless another datum is specified. Approximate times are identified. Expand an event to see its evidence.
About 13:19Departure from New Orleans
The scheduled passenger flight departed for Kennedy with 116 passengers and eight crew.
Flight 66 heard Eastern 902's report and confirmed it to the final vector controller. It did not receive Flying Tiger 161's warning and recommendation to change runways.
The NTSB found no evidence of a preimpact failure of the structure, flight instruments, controls or engines. All three engines were operating normally until impact.
They were nonfrangible structures. Their impacts caused severe damage to the left wing and fuselage as the aircraft crossed the approach lighting system.
The simulator wind models required assumptions about winds above and below the reconstructed flightpath. The pilots anticipated the encounter, and attempts to reproduce surprise were unsuccessful.
Actions and recommendations are described at the dates of the cited records. A recommendation is not proof of implementation.
Detecting dangerous winds
The NTSB recommended equipment to detect wind variations near active runway ends, together with interim surface wind measurements and lighted windsocks.
The Board recommended timely thunderstorm information for controllers and pilots, and runway selection criteria that considered storm location and severity.
The Board recommended research into preserving vertical guidance during the change from instrument to visual flight, and airborne devices to alert pilots during windshear encounters.
The FAA criteria issued before the crash required frangible structures for future approach light installations. The report said a retrofit program would be considered if funding became available.
Matching dossiers for Approach light structures: No other dossier in this collection.
Read alongside
These comparisons connect specific research topics. Shared circumstances do not establish identical causes. Evidence from both investigations is provided below.
Eastern 66 and Delta 191 both encountered severe thunderstorm winds at low altitude on final approach. Eastern's investigation identified a downdraft and decreasing headwind, with continued use of the hazardous approach as a contributing factor. Delta's investigation identified a microburst encounter and included the decision to continue toward visible lightning, gaps in avoidance and escape training, and the lack of timely windshear hazard information among the probable causes.
Explore the records below for additional context. They open on the publisher's site and are not all included in our report search. Source reuse terms apply.
Kevin Murphy writes the incident accounts on AirlineIncidents.com, out of a long interest in aviation safety and the lessons each investigation leaves behind.
Prepared by
AirlineIncidents editorial desk
Report citation check
Independent expert review
Not completed
Citation checks compare supporting text with the indexed report pages. They do not constitute independent expert review. How these pages are prepared.
Revision history
: NTSB final report NTSB/AAR-76/08, adopted 12 March 1976.
.jpg){kind=link}