TRACE THE
FAILURE
AVIATION ENGINEERING

Aloha Flight 243.
The roof tore away at 24,000 feet.

The dramatic opening in the fuselage began with damage accumulating along a seam that inspections were meant to protect.

Follow the evidence
Illustrative airliner cruising above clouds before any fuselage damage
Still from an AI-generated episode shot of an airliner in flight. Not a photograph of Flight 243 or the actual accident aircraft.
THE COMPANION FILMThe Plane That Lost Its Roof at 24,000 Feet Watch episode 3:25

Why did Flight 243’s fuselage fail?

Aloha Airlines Flight 243 suffered a major fuselage failure because disbonding and fatigue damage in a longitudinal skin joint went undetected. The NTSB identified the airline’s maintenance program as the probable-cause failure, with contributing problems in management, regulatory oversight and the response to the joint’s known vulnerabilities. NTSB executive summary.

On April 28, 1988, the Boeing 737-200 was flying from Hilo toward Honolulu when a large section of its upper fuselage separated at 24,000 feet. The crew descended and landed at Kahului, Maui. One flight attendant lost her life. The failure began along a seam joining overlapping sheets of the pressure cabin. NTSB summary.

Watch the explanation

Follow the lap joint from bonded skin to rivet loading and linked fatigue cracks.

What a lap joint was supposed to do

A lap joint connects overlapping skin panels. On early 737s, an adhesive bond and rows of rivets joined those panels. The adhesive was intended to transfer the load associated with cabin pressurization across the overlap. FAA, Lap Splice Design.

Each pressurization cycle loads the fuselage. If the adhesive bond deteriorates or separates, the rivets carry loads the adhesive was meant to distribute. Countersunk rivet holes in this early design left thin, highly stressed edges in the outer sheet, making those locations vulnerable to fatigue cracking. NTSB, section 2.2; FAA, Widespread Fatigue Damage.

Corrosion and disbonding could reinforce one another. The FAA describes difficulties with the cold-bond manufacturing process. In service, moisture could enter disbonded areas and cause further deterioration. Repeated loading then acted on the damaged joint, while inspections failed to find the damage in time.

Why many small cracks mattered

Fatigue damage can develop at several neighboring rivet holes. Each crack may initially be small, yet the remaining material between cracks becomes less capable of carrying load. When cracks join, a long fracture can form much more abruptly than an inspection strategy built around one growing crack might suggest. NTSB analysis and findings; FAA, Widespread Fatigue Damage.

Features designed to stop one local crack may no longer work as intended if neighboring joints and crack-arrest features have also deteriorated. The NTSB had to reconstruct how the failure began from the remaining airframe and its service history: the initiating portion was not recovered. Investigators could not directly examine the first crack opening. NTSB, section 2.2.

The maintenance lesson

The NTSB examined how the airline planned, performed and supervised inspections, what the FAA required and monitored, and whether known cold-bond problems had been fully corrected. These maintenance and oversight failures allowed the fatigue damage to go undetected. NTSB executive summary and conclusions.

After the accident, inspections of aging aircraft intensified and corrective actions followed. Inspection methods, intervals and repairs have to account for the damage an aircraft can develop in service, including fatigue at multiple locations after a joint loses its intended load-sharing behavior. NTSB, printed p. 46 / PDF p. 53.

The source record

NTSB Aircraft Accident Report AAR-89/03

June 14, 1989 · Executive summary, printed p. v / PDF p. 5; analysis, printed pp. 47–49 / PDF pp. 54–56; findings and probable cause, printed pp. 71–73 / PDF pp. 78–80.

FAA Lessons Learned: Aloha Airlines Flight 243

Accident Overview; Lap Splice Design; Widespread Fatigue Damage; service history and regulatory response.

Research checked September 21, 2026. Illustrative episode images are not primary evidence.