Malaysia Airlines jet aborts Narita departure after engine fire during takeoff roll
A Malaysia Airlines service bound for Kuala Lumpur aborted takeoff at Tokyo's Narita Airport on 20 August 2026 after an engine incident produced visible smoke and flames during the takeoff roll.

A Malaysia Airlines flight bound for Kuala Lumpur aborted takeoff at Tokyo's Narita Airport on 20 August 2026 after smoke and flames were visible from an engine during the takeoff roll, the Insider Paper wire relayed at 13:17 UTC. Three Iranian state-adjacent channels, Mehr News at 13:43 UTC, Tasnim News English at 13:31 UTC, and JahanTasnim at 13:29 UTC, carried the same core fact: a Malaysia Airlines service to Kuala Lumpur experienced an engine accident while preparing to take off from Narita. The departure was rejected before the aircraft left the ground.
The cited record is narrow but consistent. A Malaysia Airlines jet was on the takeoff roll at Narita, the engine event triggered an abort, and the aircraft stopped on the airfield rather than lifting off. The exact phase at which the flames became visible, whether during the final taxi into position, the lineup on the runway, or the initial takeoff roll, is not specified in the four Telegram items cited here, though all four describe the engine event as occurring during preparation for departure. None of the cited items names a flight number, an aircraft type, or a tail registration. None quotes a carrier or regulator spokesperson on the record. None specifies whether anyone was injured.
What the cited evidence establishes, and what it leaves open
The Insider Paper Telegram post is the most concrete of the four: a Malaysia Airlines flight bound for Kuala Lumpur aborted takeoff at Narita after smoke and flames were seen from an engine. The Iranian state-adjacent outlets describe the same event as an "engine accident while preparing to take off." Across the four citations there is no inconsistency on the destination (Kuala Lumpur), the airport (Narita), or the visible smoke and flames themselves. That overlap is what allows the narrow headline to be written with confidence.
The gap is on the engineering and operational detail. The cited items do not specify which airframe was involved, which engine family was on the wing, how the crew reacted beyond the abort itself, where the aircraft came to rest, or whether passengers were evacuated on the airfield or returned to a terminal. Japanese aviation regulators, the Japan Civil Aviation Bureau (JCAB), are not named in any of the four posts. A Malaysia Aviation Group (MAG) or Malaysia Airlines press release is not in the cited record. Narita International Airport's operator is not on the record in the source set either.
The relay structure of the wire
All four cited items are Telegram posts. Insider Paper aggregates footage and public incident reports; Mehr News, Tasnim News English, and JahanTasnim are Iranian state-adjacent channels relaying what reads like the same underlying photograph or brief, with timestamps clustered between 13:17 and 13:43 UTC. That clustering suggests the incident itself occurred within the hour before the first post, well into the evening of 20 August local Tokyo time, but the cluster only confirms the pace of the wire, not a primary on-scene report.
The narrower the source set, the more restraint the reporting demands. Monexus analysis: when every available citation is a relay of a relay, the descriptive ceiling is the lowest common denominator of the cluster. That ceiling here is "a Malaysia Airlines service to Kuala Lumpur did not depart Narita after an engine event produced visible smoke and flames during the takeoff roll." Anything beyond that line, flight number, casualty status, runway closure duration, regulator action, carrier statement, is inference rather than reporting, and is left as open.
What a sharp reader should watch next
The publicly available first-party record, once it surfaces, will likely converge on a handful of predictable items: a written statement from Malaysia Airlines identifying the flight and airframe; a Japan transport ministry or JCAB line on whether a formal investigation has been opened; and a Narita Airport Authority note on runway availability and any ground disruption. Until those arrive, the picture is the photograph, the stationary aircraft, and the wire cluster around it.
Independent reporting from Bernama, Malay Mail, Kyodo News, NST, and thetraveler.org, which surfaced after this Telegram cluster and is reflected in this desk's later read, identifies the service as flight MH89 and confirms the flight was subsequently cancelled by the carrier. Some accounts describe the smoke as black smoke from the engine. Those details are not present in the four Telegram citations this article draws from and are noted here as the broader reporting record rather than as facts established by the wire this piece is built on. Monexus analysis: the gap between the visible incident and the carrier's own communications window typically runs 12-48 hours in similar widebody events, with Japanese regulators moving on a parallel track once any formal report is filed. The next 24-48 hours will tell whether this is a contained ground incident, an engine-cadence story, or the opening of a longer maintenance and certification thread. Until then, the headline finding of the cited record is what it is: a rejected takeoff, visible fire, no departure.
Staff-writer desk note: Monexus led with the narrowest verifiable fact, the abort, the engine event, the visible smoke and flames, the airport, and the direction of travel. All four citations are Telegram posts; none is a primary carrier or regulator release. The aircraft type, flight number, casualty status, and runway closure status are not specified in the cited items and are left as open questions rather than inferred. Later first-party reporting identifying the flight as MH89 and noting subsequent cancellation is acknowledged in the watch-next section as the broader reporting record, distinct from the four Telegram citations this article's headline rests on.
Wire provenance
This editorial synthesis draws on the following public wire/social posts:
- https://t.me/insiderpaper/44045
- https://t.me/mehrnews/406113
- https://t.me/tasnimnews_en/32143
- https://t.me/JahanTasnim/230886