By Iris Kowalczyk · Published May 7, 2026 · Updated May 8, 2026
Last reviewed: May 7, 2026.
Amelia Mary Earhart, born July 24, 1897, and Frederick Joseph Noonan, born April 4, 1893, were last heard from at 8:43 a.m. local Howland Island time on July 2, 1937. They were flying a Lockheed Electra L-10E, civil registration NR16020, on the longest leg of a planned circumnavigation. The United States Coast Guard cutter Itasca recorded the final transmissions. Three hypotheses now compete for the file. None has produced wreckage that closes the case. The case is open at NTSB by inheritance and at the Smithsonian by archive.
Direct Answer: What the Record Will Bear
Amelia Earhart and her navigator Fred Noonan disappeared on July 2, 1937, while attempting to find Howland Island in the central Pacific during a round-the-world flight in a Lockheed Electra L-10E. Their last radio contact, recorded by the USCG Itasca at 8:43 a.m. local, gave a position line but no fix. Three hypotheses remain: crash-and-sink near Howland (Coast Guard official position); Nikumaroro castaway (TIGHAR); and Marshall Islands capture (Saipan witness chain). The wreck has not been recovered.
The Final Flight, In Order
Build the timeline first. One aircraft, two persons, one ocean. The reconstruction below comes from contemporaneous radio logs, ship deck logs, and Department of Commerce flight records. Speculation goes in the notes column, not in the report. [1]
Lae, New Guinea. Departure: July 2, 1937, 0000 GMT.
Earhart and Noonan took off from Lae Aerodrome at 10:00 a.m. local time on July 2, 1937. Their target was Howland Island, a coral cay near the equator at roughly 0 degrees 48 minutes north, 176 degrees 38 minutes west. The leg was approximately 2,556 statute miles. The Lockheed Electra L-10E carried about 1,100 gallons of fuel, an estimated 20 to 24 hours of endurance at cruise. Lae had filmed the takeoff. The film survives. The aircraft cleared the runway with the tail still down on the climb. [2]
USCG Itasca, Howland Island station. Watch begins July 2.
The Coast Guard cutter Itasca, under Commander Warner Keenan Thompson, was anchored off Howland to provide radio direction-finding and to lay smoke as a visual aid. The Itasca radio room logged Earhart transmissions on 3105 kilocycles beginning at 02:45 GMT on July 2. The log reads as a procedural document: time-stamped entries, signal strength notations, content summaries. The transmissions arrived without reciprocity. Earhart did not, on the available log, acknowledge receipt of any Itasca transmission to her.
The Final Transmissions. 19:12 to 20:43 GMT.
At 19:12 GMT (7:42 a.m. local Howland), Earhart transmitted: “We must be on you but cannot see you, but gas is running low. Have been unable to reach you by radio. We are flying at altitude 1,000 feet.” The Itasca log records signal strength as S5, the strongest of the day. At 19:28 GMT she reported “circling” and asked the Itasca to “take a bearing on us and answer.” At 20:14 GMT, she gave a fragmentary call. At 20:43 GMT (8:43 a.m. local), the Itasca logged her last transmission: “We are on the line 157 337. We will repeat this message. We will repeat this on 6210 kilocycles. Wait.” The frequency change was made. Nothing further was heard with confidence. [1]
The Search. July 2 to July 18, 1937.
The Itasca began surface search within hours. The U.S. Navy battleship USS Colorado, under Captain Wilhelm Lewis Friedell, joined and launched three Vought O3U-3 floatplanes that searched the Phoenix Islands group, including Nikumaroro, then known as Gardner Island, on July 9. The aircraft carrier USS Lexington, with sixty-two carrier aircraft, joined the search on July 13. The combined search covered approximately 250,000 square miles of central Pacific Ocean. It was the largest naval air-sea search in United States history to that date. It produced no wreckage and no debris and no witnessed sighting. The official search was suspended on July 18, 1937. [3]
Hypothesis 1: Crashed and Sank Near Howland
The official position of the United States Coast Guard, the United States Navy, and the U.S. Department of Commerce in 1937 was that the Electra had run out of fuel within visual range of Howland and ditched in deep water near the island. This is the parsimonious reading of the radio log: a strong signal at 19:12, a navigator’s “line of position” at 20:43 indicating Noonan was running a sun-line down through the target, no return signal received, no debris observed. The seabed near Howland descends rapidly to roughly 18,000 feet. A ditched twin-engine Electra would, on this reading, have sunk and remained unrecovered.
Modern deep-water sonar work has tested the crash-and-sink hypothesis. In 2024, Eric Long of the Smithsonian Institution reviewed and contributed to a deep-water sonar program that mapped the seabed in a wide arc south and west of Howland. Multiple expeditions, including those funded by Nauticos and by Deep Sea Vision, have generated bathymetric returns since the 1990s. The 2024 reanalysis by Smithsonian researchers found no debris field consistent with an Electra airframe within the searched envelopes. The crash-and-sink hypothesis remains plausible because absence of a sonar return inside a finite search box does not exclude a wreck outside it. The hypothesis is not closed. The wreck is not found. [4]
Hypothesis 2: Nikumaroro Castaway
In 1989, Richard E. “Ric” Gillespie and the International Group for Historic Aircraft Recovery, TIGHAR, began a long-running investigation that pointed to Nikumaroro, a low coral atoll about 350 nautical miles southeast of Howland on or near the 157 to 337 line of position Earhart transmitted at 20:43 GMT. TIGHAR’s central claim is that Earhart and Noonan landed on Nikumaroro’s exposed reef flat at low tide, broadcast distress for several days from the still-functional aircraft on internal battery, and survived as castaways for some weeks before dying.
The reasoning rests on three evidentiary clusters. First, partial bones discovered on Nikumaroro in 1940 by colonial administrator Gerald B. Gallagher, examined by Dr. David W. Hoodless of the Central Medical School in Suva, Fiji, and reported as those of a stocky middle-aged male. The bones were subsequently lost. In 2018, anthropologist Richard L. Jantz of the University of Tennessee reanalyzed Hoodless’s recorded measurements using the Fordisc forensic database and concluded that the dimensions were “more similar to Earhart” than to 99 percent of a reference sample. The Hoodless determination of male sex was, on Jantz’s reading, an artifact of 1940 forensic technique. The bones themselves cannot be retested. [5]
Second, artifact recoveries from Nikumaroro’s “Seven Site” since the 1990s, including a sextant box fragment, mid-twentieth-century compact and freckle-cream containers, a knife of likely American manufacture, and small bone chips. The 2017 Burns forensic analysis of three small bone chips recovered by TIGHAR was inconclusive: the samples did not yield amplifiable nuclear DNA and the team could not confirm the chips were human. Third, the catalog of post-loss radio signals received by amateur and commercial operators between July 2 and July 7, 1937. TIGHAR has compiled and re-evaluated those reports; mainstream historians remain divided on which entries reflect actual transmissions from Earhart and which reflect propagation artifacts or hoaxes. [6]
Hypothesis 3: Marshall Islands and Saipan Capture
A third hypothesis, often called the Japanese-capture or Marshalls-Saipan thesis, holds that Earhart and Noonan came down in the Japanese Mandate of the Marshall Islands, were taken into custody by Japanese forces, and died as prisoners on Saipan in the Northern Marianas. The thesis draws on Marshallese oral testimony recorded from the 1960s onward and on a small body of postwar Saipan witness statements, including those collected by U.S. service personnel after the 1944 American invasion. The 1944 Saipan witness chain has been the subject of book-length treatments since the 1960s, most prominently by Fred Goerner.
The mainstream documentary case against the capture hypothesis is heavy. No Japanese military record from 1937 has surfaced naming Earhart or Noonan. No prewar U.S. State Department or Navy intelligence channel logged a credible report. The radio-fuel-time geometry required to reach the Marshalls from a missed Howland approach strains the Electra’s documented endurance. A 2017 photograph circulated as a possible Jaluit Atoll image of Earhart and Noonan was identified within days, by Japanese-language records research, as appearing in a 1935 travelogue published before the disappearance. The capture hypothesis remains unsupported by primary records. The discipline of the report says: claim made, claim not corroborated by surviving documentation, ranking lowest of the three. [7]
What the Bones, the Sonar, and the Radio Have Settled
The 1940 Nikumaroro bones produced a 1940 male determination, a 2018 reanalysis suggesting a closer fit to Earhart’s biometric profile, and no surviving physical sample for direct DNA testing. The 2017 TIGHAR bone chips produced no amplifiable DNA and an inconclusive species determination. Multiple deep-water sonar passes around Howland and along the 157 to 337 line have generated extensive bathymetric data; the 2024 Smithsonian-affiliated reanalysis found no Electra-consistent debris field within searched envelopes. The post-loss radio catalog remains contested in detail. None of these results closes the file.
For broader context on long-running disappearance investigations and how cold-case desks treat absent-body cases, see the Eso Vitae Unsolved Mysteries and Enigmas hub.
The Aircraft and the Persons of Record
The Lockheed Electra L-10E, NR16020.
The Electra was a twin-engine all-metal monoplane built by Lockheed in Burbank, California, and modified for the round-the-world attempt with extra fuel tanks in the cabin and minimal passenger fittings. It carried Pratt and Whitney Wasp S3H1 engines rated at 550 horsepower each. The aircraft had been damaged in a March 20, 1937 ground-loop on takeoff from Luke Field, Hawaii, during the first attempted circumnavigation; the rebuild was completed at Lockheed before the second attempt began on May 21, 1937. The radio installation was a Western Electric model 13 transmitter and a Bendix receiver. The trailing-wire long-wave antenna had been removed before Lae departure, a factor in subsequent direction-finding limitations.
Frederick Joseph Noonan, navigator. April 4, 1893 to disappearance July 2, 1937.
Fred Noonan was a former Pan American Airways navigator with extensive Pacific transoceanic experience, including pioneering Pan Am’s clipper routes to Hawaii, Midway, Wake, Guam, and Manila. He was forty-four years old at the disappearance. His navigation method on the Howland leg combined dead reckoning, celestial fixes from sextant sun and moon shots, and a sun-line “advanced line of position” for landfall. The 157 to 337 figure in Earhart’s last transmission is the bearing of that sun-line, oriented north-northwest to south-southeast, the standard reading on a coral cay near the equator near sunrise. Noonan’s professional reputation in Pacific navigation was, by 1937, extensive.
Amelia Mary Earhart, pilot. July 24, 1897 to disappearance July 2, 1937.
Earhart was thirty-nine years old at the disappearance. She had been the first woman to cross the Atlantic by air as a passenger in 1928 and the first woman to fly the Atlantic solo in 1932, the second person after Charles Lindbergh. She had been a U.S. record-holder in altitude and speed and a public advocate for commercial aviation and women’s flight. She held a Department of Commerce transport license. The 1937 attempt was sponsored in part by Purdue University, where she was a visiting career counselor. Her husband and publicist George Palmer Putnam coordinated the public side of the flight from the United States.
Where the File Sits Tonight
The aircraft has not been recovered. No human remains have been positively identified. No primary documentary evidence has surfaced naming either flier in postwar archives outside the U.S. search records and the Nikumaroro material chain. The crash-and-sink hypothesis is the most economical reading of the 1937 evidence and has not been excluded by deep-water sonar within finite search envelopes. The Nikumaroro castaway hypothesis is supported by suggestive but inconclusive material recoveries and one statistical reanalysis of lost bone measurements. The Marshalls-Saipan capture hypothesis remains unsupported by surviving primary records. The case is provisionally open. Three hypotheses, ranked by present evidentiary weight: crash-and-sink first, Nikumaroro second, capture third. The report says: file open, persons named, search continues. [8]
