Here’s the thing about flying. Every commercial aircraft in the world has two pilots up front. There’s a good reason for that. If something goes wrong with one of them, the other one lands the plane. That is the whole point of having two.
Which means, when you really think about it, the aviation industry has to be extremely careful about anything that could take out both pilots at once. Not the plane. Not one of the pilots. Both of them, at the same time, in a way that leaves nobody up front who can fly.
One of the things that could do that, and the specific thing this rule exists to prevent, is the meal on the tray in front of them.
What happened on the flight from Anchorage
The morning of 3 February 1975, a Japan Air Lines Boeing 747 lifted off from Anchorage International Airport bound for Copenhagen. It had originally departed Tokyo, was headed onward to Paris, and was carrying 344 passengers and 20 crew. Most of the passengers were Japanese sales representatives from Coca-Cola, on a company-paid trip to Europe. Anchorage was the second stopover on a long-haul route. The passengers had breakfast about an hour before the aircraft approached Copenhagen.
What was on the tray was ham and omelette. It had been prepared the previous day at International Inflight Catering, a Japan Air Lines subsidiary in Anchorage. The ham had been held at room temperature for around fourteen hours in the kitchen, then at 10 degrees Celsius, or 50 Fahrenheit, for another fourteen and a half hours, before finally being loaded onto the aircraft ovens and warmed for service.
According to the encyclopaedic record of the incident maintained on Wikipedia, drawing on contemporaneous New York Times reporting and the subsequent Lancet investigation, one of the three cooks who had prepared those omelettes at International Inflight Catering had infected lesions on the index and middle fingers of his right hand. He had bandaged them and gone to work. He had not told his manager. The lesions were carrying Staphylococcus aureus, and the specific ham he had handled with those hands was, as it sat at room temperature for the better part of a day, developing enough staphylococcal toxin to make several hundred people extremely ill.
About an hour after breakfast, roughly 143 of the passengers began vomiting. One flight attendant went down with them. All of them were hospitalised in Copenhagen. Thirty of them required what the reporting at the time called intensive care. The aircraft landed safely. Everyone survived.
But the specific reason everyone survived, on the accumulated evidence of the investigation, was a single accident of timing.
If you find this interesting, you’ll love this video on the hidden effects of long-haul flying.
The pilots ordered steak
The pilots had come off a very long shift. Their biological clocks were still on Alaska time, not European time. So when the breakfast trays came around, they didn’t feel like breakfast food. According to the Wikipedia record of the Copenhagen investigation, they had ordered steaks instead of the ham omelettes. That was the only reason they were still functional when the sickness swept through the cabin.
The investigation into the Anchorage catering facility was led by Dr Mickey Eisenberg, a United States Public Health Service officer with the Alaska State Health Department. He and his colleagues traced the outbreak to the cook, catalogued the specific bacterial phage types and antibiotic resistances that linked the cook’s finger lesions to the ham on board, and published their full findings in the medical journal The Lancet on 27 September 1975, in a paper titled “Staphylococcal Food Poisoning Aboard a Commercial Aircraft”.
The paper’s abstract, which now sits archived in the University of California’s eScholarship system, includes a single line of policy recommendation that most people who read it at the time did not expect to become an industry standard. It reads: “To ensure against a common foodborne illness incapacitating the entire flight crew, cockpit crew members should eat different meals prepared by different cooks.”
That sentence, buried at the end of an epidemiology paper about ham omelettes, is where the modern rule comes from.
Seven years later, over the Atlantic
In August 1982, roughly seven years after the JAL incident, an Overseas National Airways flight was heading from Lisbon to Boston. Around thirty minutes out from Logan Airport, ten of the twelve crew members on board began to feel very sick very quickly. Among them were the captain, the first officer, and the flight engineer.
They had all eaten the tapioca pudding.
The pudding, which was later traced to a Lisbon catering source, made at least six passengers on a separate Capitol Airlines flight from Lisbon sick three days later. The specific incident was reported in the New York Times two years afterwards, when the paper ran a piece on the state of commercial aviation food safety. A spokesman for the Massachusetts Port Authority summarised the aviation-safety implication with the kind of understatement that makes airline incidents sound perfectly manageable: the crew all eat different things, he said, but the one thing they had in common was tapioca pudding.
The plane landed safely. Nobody died. But the specific combination of “captain, first officer, and flight engineer all incapacitated at once from the same dessert” is the exact scenario Eisenberg’s 1975 recommendation had been written to prevent. In the seven years between the two incidents, most airlines had not implemented it.
What the rule actually looks like
The rule today is not, and has never been, a formal regulation from the Federal Aviation Administration. According to an aviation safety management industry analysis published in March 2026 by SafeJets, a firm that specialises in Safety Management System frameworks for commercial and business aviation, the specific rule is codified inside each individual airline’s Operations Manual, or within the Air Operator Certificate safety documentation, or as a Standard Operating Procedure within the Safety Management System that ICAO and national regulators require every commercial operator to maintain.
What that means in practice is a set of specific operational controls. The catering supplier prepares the captain’s meal and the first officer’s meal in separate batches, sealed and labelled, and often sourced from different ingredient suppliers. Before pushback, the two pilots confirm to each other in the cockpit that they have not chosen the same meal. If they have, one of them changes. During the flight, they eat at different times, so that if either meal is going to produce symptoms, they will show up in one pilot before the other, giving the aircraft enough time to divert.
Some airlines allow the captain to eat from the first-class galley while the first officer eats from the business-class galley, which by definition means the meals were prepared and stored separately. Some go further and require crew meals to be prepared in an entirely different kitchen from passenger meals, then delivered to the aircraft in sealed containers that only the crew opens. Some allow pilots to bring their own food from home, which many pilots quietly prefer.
And when something does go wrong, the Standard Operating Procedure defines what happens next. If both pilots have inadvertently consumed the same meal, and either one begins showing symptoms, the flight is diverted. Not paused, not monitored, not maybe. Diverted.
What the rule is really about
The specific philosophy of aviation safety that this rule sits inside is called redundancy. It is the principle that no single failure, of any kind, should be able to bring down an aircraft. Every critical system on a modern airliner has a backup. Two engines. Two hydraulic systems. Two electrical buses. Two pilots. Two of almost everything, so that if any one of them fails, the other one carries the aircraft to a safe landing.
The uncomfortable observation the JAL and Overseas National Airways incidents forced the industry to make is that the two pilots up front are not, by default, actually independent. They fly the same aircraft. They breathe the same cabin air. They talk to the same passengers. And, until Eisenberg’s 1975 recommendation was slowly adopted through the 1980s and 1990s, they routinely ate the same meal from the same tray. Which meant that in the specific category of hazards involving anything at all in that shared environment, the redundancy of having two pilots collapsed to a redundancy of one.
The split-meal rule is an attempt to un-collapse it. The two pilots share almost everything about their working environment, but they do not share their food. The bacteria on the cook’s finger in Anchorage that morning in 1975, or the pudding batch in the Lisbon kitchen in August 1982, do not get to affect both pilots at the same time. One of them gets sick. The other one lands the plane.
What makes the rule genuinely unsettling, on reflection, is not the specific scenario it prevents. It is what the specific scenario implies about the ordinary operational picture that everyone accepts without thinking. Every commercial flight is being carried, right now, over an ocean or across a continent, by two people who share the same tiny volume of air, the same tiny volume of physical space, and the same set of environmental exposures for hours at a time. The industry has taken decades to identify and separate out, one by one, the specific shared exposures that might take out both of them at once.
The food is one of the ones they’ve managed to separate. Most of the others, they have not.
The rule about the meal is not there because the food is unsafe. It is there because the flight, on the accumulated evidence of six decades of aviation medicine, is not quite as redundant as it looks.
