Malaysia Airlines Flight MH370

The first hour is documented by air traffic control recordings, the aircraft's ACARS reports (the last at 01:07), and civil and military radar⁠[1]. The turn-back began within two minutes of the transponder stopping and took the aircraft through a sequence of turns and a route along airway waypoints that the investigators judged was flown by hand or with the autopilot deliberately directed, not by a system failure⁠[1]. The investigation found no fault in the aircraft or its maintenance, nothing in the cargo, nothing in the passengers' backgrounds, and no evidence that resolved the crew's state of mind; the captain's home flight simulator held a deleted route into the southern Indian Ocean, which the report noted without drawing a conclusion⁠[1].

The six hours after radar are documented only by the Inmarsat log. The satellite data unit's handshakes at 02:25, 03:41, 04:41, 05:41, 06:41, 08:11 and 08:19 each carry a burst timing offset (BTO) and a burst frequency offset (BFO). The BTO fixes a ring of positions on the Earth for each handshake; the BFO, corrected for the satellite's own motion, constrains the aircraft's velocity component toward the satellite and distinguishes a northern from a southern route⁠[2]. Inmarsat's engineers validated the method against other aircraft flying known routes that night⁠[2]. Malaysia released the log, 47 pages, on 27 May 2014⁠[9]. The last two signals, 08:11 and 08:19, are read as the aircraft running out of fuel on one engine and then the other, the auxiliary power unit starting and the SATCOM rebooting, which places the end of flight within a few tens of kilometres of the seventh arc⁠[4]⁠[3].

The debris is the physical record. The flaperon found on Réunion on 29 July 2015 was identified by French investigators from its internal serial plates; the outboard flap from Pemba carried a part number and Malaysia Airlines' stencilling; a third piece was likewise matched⁠[5]. Examination of the flap and flaperon found that the flaps were not extended at the time of separation, which argues against a controlled ditching with flaps down⁠[5]. Marine growth on the flaperon and the timing of the finds, compared with drifting buoys and with replica flaperons tracked by CSIRO, were used to back-calculate where the pieces entered the water⁠[6].

The surface search in March and April 2014 began in the South China Sea, moved to the Strait of Malacca when the radar track was understood, and then, once the Inmarsat analysis was accepted, to the southern Indian Ocean, where it found nothing attributable to the aircraft⁠[1]. The underwater search was led by the Australian Transport Safety Bureau on Malaysia's behalf with Chinese participation. The search zone was set by a Bayesian analysis of the satellite data by Australia's Defence Science and Technology Group, which produced a probability distribution along the arc and a 120,000 km² priority area centred near 38°S⁠[3]. Towed and autonomous sonar surveyed it from October 2014 until the search was suspended on 17 January 2017; the area was cleared to a high standard of confidence, and the ATSB's final report concluded that the aircraft was not in it⁠[4].

Two lines of evidence then moved the estimate north. CSIRO's drift modelling of the debris finds, using actual flaperon replicas at sea, found that an origin near 35°S was most consistent with the arrival dates and with the absence of debris on Australian shores⁠[6]. A review of the final handshakes and of the aircraft's likely flight dynamics at fuel exhaustion suggested the aircraft came down close to the arc and that the unsearched stretch between about 32.5°S and 36°S, some 25,000 km², was the most likely remaining area⁠[4]. Ocean Infinity's vessel Seabed Constructor searched that stretch and more, about 112,000 km², with a fleet of autonomous vehicles from January to June 2018, on no-find, no-fee terms, and found nothing⁠[1]⁠[7].

Between 2018 and 2024 independent analysts proposed narrower targets from refinements of the satellite analysis, from drift studies, and from amateur radio (WSPR) propagation records whose use for aircraft tracking is disputed. Ocean Infinity proposed a new search in 2024; the Malaysian cabinet approved it in December 2024 and agreed terms in March 2025: $70 million payable only on finding the wreckage, for a 15,000 km² area⁠[7]. The vessel worked the site from 25 to 28 March 2025 before withdrawing, and again from 31 December 2025 to 23 January 2026, surveying 7,571 km² before weather and a reassignment ended the campaign. In March 2026 Malaysia's Air Accident Investigation Bureau stated that the search "has not yielded any findings that confirm the location of the aircraft wreckage" and gave no date for resumption; the contract runs to June 2026⁠[8].

Deliberate diversion by someone on the flight deck is the reading most analysts hold. It accounts for the timing of the transponder loss at the handover between control centres, the hand-flown turn, the route along the Malaysia–Thailand border that avoided both countries' attention, the silence of every communication system except the one nobody thought to turn off, and the long southern leg⁠[1]. The Malaysian investigation did not name a perpetrator and said it could not exclude third-party interference; it found nothing in the crew's records that showed intent, and the captain's simulator route is suggestive and no more⁠[1]. The reading divides on the end of flight: an unpiloted descent at fuel exhaustion, which the final BFO values and the flap evidence support, or a controlled glide, which would put the wreckage farther from the arc than any search has looked⁠[4]⁠[5].

Accident readings propose a fire, depressurisation or electrical failure that incapacitated the crew after a turn toward a diversion airport, with the aircraft flying on autopilot until the fuel ran out. Against them: the turn-back was hand-flown through several headings, the loss of the transponder and ACARS preceded any sign of trouble, and no system failure explains why the SATCOM came back after an hour⁠[1]. Hijacking by a passenger is not excluded by evidence but is unsupported by any claim, background or trace⁠[1]. Readings that place the aircraft elsewhere than the southern arc, including a northern route, Diego Garcia, or a shoot-down over the South China Sea, are contradicted by the Inmarsat data, the debris and the drift analysis⁠[2]⁠[6].

Finding the wreckage. The flight data and cockpit voice recorders, if recoverable after a decade at depth, would show who did what; the wreckage pattern would show whether the aircraft came down fast or was ditched. The remaining question is where to look. The 2025–2026 campaign covered half its target; the rest of that area, the stretches of the arc beyond 25 nautical miles that a glide could reach, and the latitudes north of 33°S are the candidates⁠[4]⁠[8]. Short of the wreckage, a re-examination of the satellite data with the methods and the drift knowledge now available, published with the raw log, would narrow the field; and any further debris with a recorded find date and location tightens the drift back-calculation⁠[6].