In March 1968, a Soviet ballistic-missile submarine named K-129 left Kamchatka for a routine Pacific patrol and never came home. The Soviets searched for months and found nothing. The United States found the wreck about 1,560 miles northwest of Hawaii — nearly three miles down, on a black abyssal plain no salvage ship had ever worked. Six years later, the CIA tried to steal the submarine off the seafloor with a claw the size of a building, under a cover story so loud it almost worked.
That effort was Project Azorian. The ship was the Hughes Glomar Explorer. And the phrase the operation left behind — “we can neither confirm nor deny” — still answers FOIA requests today.
A Golf II, lost without a wake
K-129 was a Golf II-class diesel-electric boat, NATO’s name for the Soviet Project 629A. She carried three SS-N-4 Sark ballistic missiles, each with a one-megaton warhead, plus torpedoes. Her peacetime station sat northeast of Hawaii. If war had broken out, those missiles were aimed at the U.S. West Coast.
On March 1, 1968, she sailed from Petropavlovsk. Sometime in mid-March she suffered a catastrophic accident and sank with her entire crew. The CIA’s later declassified history does not say what caused the loss. What mattered to Washington was that the Soviets never found the wreck — and that U.S. underwater listening systems had a fix.
By August 1968, the Navy had photographed the site. The submarine lay roughly 16,500 feet down. Raising anything that heavy from that depth had never been done. The deepest prior U.S. submarine salvage, USS Squalus in 1939, had been from 245 feet.
Howard Hughes, manganese nodules, and a moon pool
In April 1969 the White House gave the CIA the lead, with Defense Department support. The only plan that looked feasible was a surface ship with a massive claw — officially a “capture vehicle,” nicknamed Clementine — lowered on a string of steel pipe sections, the way an oil derrick adds drill pipe.
The cover had to be better than quiet. Billionaire Howard Hughes lent his name and eccentricity. The public story: the ship would mine manganese nodules from the deep ocean floor. The real ship, built at Sun Shipbuilding in Chester, Pennsylvania, and finished over several years, looked like a mining vessel on purpose. It had a derrick, pipe-handling gear, docking legs, and a huge center well called a moon pool, big enough to swallow a chunk of submarine. Doors sealed the well’s floor so the entire recovery could happen underwater, out of sight of ships, planes, and spy satellites.
The claw itself was assembled under a roof inside a barge, HMB-1, then floated out and transferred into the Glomar’s moon pool off Catalina Island — a maneuver meant to keep the hardware hidden until the last possible moment.
July 4, 1974: lift day under Soviet eyes
The Glomar Explorer sailed from Long Beach and arrived over the wreck on July 4, 1974. Recovery work ran for weeks inside a narrow North Pacific weather window. Soviet ships showed up anyway — curious about the odd “mining” vessel — and watched for days. A naval tug, SB-10, stayed close for about two weeks. From their decks, it looked like deep-sea industrial work. They never saw a submarine come up.
The lift itself was slow and mechanical. Pipe sections about 60 feet long went over the side one at a time until Clementine reached the wreck. The claw had to straddle a chosen section of hull, clamp on, then reverse the pipe string, removing sections as the load rose. CIA engineers had aimed to recover a huge forward portion — on the order of 1,750 tons and more than a hundred feet of boat.
On August 1, 1974, the winching began in earnest. Eight days later, on August 8, what remained in the claw was secured inside the moon pool. Midway up, part of the grip failed. A large section of the submarine broke free and fell back to the seafloor. The crew brought home the portion that stayed locked in the capture vehicle — about 38 feet of the forward hull, including the torpedo room.
What came up — and what didn’t
Public reporting and later books agree on the basic haul: the forward section held two nuclear torpedoes and the remains of six Soviet submariners. The CIA gave the men a formal burial at sea, filmed the ceremony, and decades later returned that film to Russia. In 1992, Director of Central Intelligence Robert Gates presented it to President Boris Yeltsin. The ship’s bell from K-129 was later returned as well, in a 1993 handover in Moscow.
What the agency most wanted — intact missiles, code material, the radio room, the sail and control spaces — was widely reported to have gone back down with the broken section, or to have been destroyed when the boat sank. Some accounts argue missile-fuel explosions had already wrecked the midships. CIA Director William Colby called Azorian an intelligence failure on the crypto and missile fronts. The agency still ranked the operation among the Cold War’s boldest technical coups: the cover held through the lift, the engineering worked at a depth that had no precedent, and something came home.
A second recovery attempt, sometimes discussed under the name Project Matador, was planned. It never sailed.
A burglary, a newspaper, and the Glomar response
Azorian’s secrecy did not die at sea. It died in Los Angeles. In June 1974, just before the Glomar sailed, thieves broke into a Hughes office and stole documents — including one tying Hughes to the CIA and the Explorer. The hunt for that paper pulled in the FBI and local police. Rumors leaked. On February 7, 1975, the Los Angeles Times connected the robbery, Hughes, the CIA, and the submarine job. Columnist Jack Anderson then put the story on national television. Journalists camped on the waterfront where the ship was being readied for round two.
The Ford administration neither confirmed nor denied. That posture hardened into FOIA practice when journalist Harriet Phillippi asked the CIA for records about its media contacts on the Glomar story. The agency refused even to say whether responsive files existed. Courts later backed the approach. Ever since, a “Glomar response” has meant: we can neither confirm nor deny.
By mid-1975 the Soviets had parked a ship on the wreck site. With the cover gone, the White House canceled further recovery. The Glomar Explorer’s brief covert career ended. After some commercial mining trials, she spent years mothballed, then returned to work as a deep-ocean drillship under other names.
Why the story still matters
Project Azorian sits at the intersection of espionage, ocean engineering, and Cold War risk. The United States spent on the order of $800 million in 1970s dollars — several billion in today’s money — to reach a single wreck three miles down, while Soviet ships watched and newspapers eventually tore the cover off. Compare that to another Cold War depth record on land: Soviet crews at the Kola Superdeep Borehole drilled more than 12 kilometers into the crust, chasing geology instead of a submarine. Both projects pushed machines into places people could not follow, and both left behind more questions than tidy victories.
Azorian also left a legal ghost. Every time an agency answers a records request with neither confirmation nor denial, the language traces back to a claw, a moon pool, and a Golf II that never finished its patrol.
The ocean kept most of K-129. The parts that rose — and the phrase that followed — are still what we have to work with.
Further reading
Project Azorian: The CIA and the Raising of the K-129
Naval historian Norman Polmar and filmmaker Michael White’s account of the Glomar Explorer mission, the lift, and what the CIA has said about the haul.
Blind Man’s Bluff: The Untold Story of American Submarine Espionage
Sherry Sontag and Christopher Drew’s classic on Cold War undersea spying — including the K-129 loss and the Glomar recovery as it looked to reporters and submariners.