At 2005 on 17 May 1987, the guided missile frigate USS Stark was told that an Iraqi aircraft had left the mainland and was heading south down the Persian Gulf. This was not alarming. Iraqi aircraft did this most nights, hunting tankers carrying Iranian oil, and they had been doing it for years.

One hundred and sixty-seven minutes later, thirty-seven American sailors were dead.

What makes the Stark worth returning to nearly forty years on is not only how it happened, though the minute-by-minute record is genuinely painful to read. It is that the most basic question about the attack has never been settled. To this day, credible sources disagree about what kind of aeroplane fired the missiles.

The war nobody was fighting

To understand why an American warship was sitting in the middle of this, you need the shape of the tanker war.

Iraq wanted to strangle Iran's oil exports, almost all of which funnelled through Kharg Island in the northern Gulf. So Iraqi aircraft flew down the western edge of the Gulf, turned east when their radar found a ship, and fired. Iran, having no Iraqi shipping to attack in return, hit tankers trading with the Gulf Arab states instead. The result was a body of water full of civilian vessels being shot at by both sides.

Rear Admiral Harold Bernsen commanded the Middle East Force at the time, and his description of Iraqi targeting practice is the single most important sentence for understanding what happened to the Stark.

Iraqi crews were not checking what they were shooting at. They found a radar return and launched. On most nights that produced a burning tanker and no headlines. On this night it produced an American frigate.

Sixty-seven minutes of small decisions

The timeline reconstructed in the US Naval Institute's Proceedings reads like a case study in how a disaster assembles itself from individually defensible choices.

Captain Glenn Brindel was told about the inbound aircraft at 2005, when it was two hundred miles away. He was reminded of it at 2015. At 2055 he asked why there was still no radar picture of it, and the combat information center switched the air-search radar to its eighty-mile setting. They acquired the aircraft at seventy miles.

At 2102, the aircraft's Cyrano air-intercept radar was detected, and for several seconds it locked onto the Stark. A radar lock is not ambiguous. It means somebody is looking specifically at you.

At 2103, the radar operator asked the tactical action officer, Lieutenant Basil Moncrief, for permission to send the standard warning. Moncrief said: "No, wait."

At 2105 the aircraft turned toward the ship at 32.5 miles, settling onto a constant bearing with decreasing range, which is the classic signature of something coming to kill you. The CIC missed it. At 2107 the first Exocet left the rail at 22.5 miles. A lookout saw the launch and reported it as a surface contact. Brindel was called and could not be found.

At 2108 the Stark finally transmitted a challenge on the military air distress frequency. The Iraqi pilot was firing his second missile as the words went out. The Phalanx close-in weapon system, the last automatic defence against exactly this threat, was placed in standby. At 2109 a lookout reported an inbound missile. The report never reached the tactical action officer.

2112

Lieutenant Art Conklin was the ship's damage control assistant. His account was published in Proceedings the following year.

The first missile came through the hull under the port bridge wing, eight feet above the waterline, at more than six hundred miles an hour. Its warhead did not detonate. This was not mercy. Because it had flown only twenty-two miles it still carried most of its fuel, and it distributed several hundred pounds of burning propellant through berthing compartments, the barbershop, the post office and the chief petty officers' quarters before coming to rest.

The resulting fire burned at over 3,500 degrees Fahrenheit, roughly double the temperature normally considered the upper limit for a shipboard fire. It melted structure.

About thirty seconds later the second missile hit eight feet forward of the first, penetrated five feet, and exploded as designed. Much of the blast vented outward through the hull, which tore a large hole in the ship and, paradoxically, saved it by not directing the energy inward.

Within minutes, close to a fifth of the crew was dead.

What happened next is the part of the story the Navy is entitled to be proud of. Gunner's Mate Third Class Mark Samples spent twelve hours spraying cooling water into the ship's missile magazine. Had it gone up, the Stark would have sunk with everyone still aboard. Seaman Mark Caouette carried on setting watertight integrity with severe burns, shrapnel wounds and one leg gone. Electronics Technician Third Class Wayne Weaver died getting other men out of the blast zone. Mess Management Specialist Second Class Francis Burke resuscitated smoke-inhalation casualties using first aid the crew had been drilled in.

They saved the ship. She was repaired and back on sea trials by August 1988.

So what was the aeroplane?

Here is where the record breaks down.

US Central Command identified the attacker as a Dassault Mirage F1, and that identification has been repeated ever since, including in the Naval Institute's own thirtieth-anniversary retrospective. The electronic warfare technicians aboard detected a Cyrano radar, which is the Mirage F1's set, so the identification was not arbitrary.

The problem is arithmetic. The F1EQ-5 variant Iraq operated could carry exactly one Exocet. The Stark was hit by two, fired from the same aircraft roughly a minute apart.

The alternative explanation is that the attacker was a Dassault Falcon 50, a three-engined business jet, modified by Iraq with a Cyrano radar in the nose, a targeting pod on the centreline and a hardpoint under each wing. That would account for both the radar signature and the second missile. In 1989 Iraqi state television broadcast footage of exactly such an aircraft, which is the strongest public evidence for the theory.

We are not in a position to settle this, and neither is anyone else outside the relevant archives. What can be said is that the official identification and the technical evidence do not agree, and that the discrepancy has never been publicly resolved. If it was a Falcon 50, then the most lethal anti-ship attack on the modern US Navy was carried out by a converted executive transport.

What it cost, and what it caused

A court of inquiry under Rear Admiral Grant Sharp found the Stark had been in a low state of defensive readiness in a high-threat area. Captain Brindel was relieved of command and recommended for court-martial along with Lieutenant Moncrief. Both ultimately received non-judicial punishment and letters of reprimand. Brindel retired early, and because he had not held the rank long enough, retired as a commander.

Bernsen's verdict on the failure was blunt, and he did not confine it to the junior officer.

There is a bleak coda. Fourteen months later, the cruiser USS Vincennes, operating in the same waters under the institutional memory of what had just happened to a ship that waited too long, shot down Iran Air Flight 655 and killed 290 civilians. Brindel himself later drew the connection, suggesting the Stark had become part of the mental calculus of every commander in the Gulf.

Bernsen was asked years afterwards what stayed with him. It was not the inquiry or the politics.

It was standing on the tarmac at Bahrain International Airport, watching thirty-six flag-draped coffins go up the ramp of a C-141 for the flight home. Thirty-six, not thirty-seven. One man went into the sea and was never recovered.

Sources: US Naval Institute (Proceedings, Naval History, oral histories of Rear Adm. Harold J. Bernsen and Adm. Frank Kelso), USNI News, Brad Peniston, No Higher Honor, Naval History and Heritage Command, The Aviation Geek Club.