1. What the 90-Second Rule Says and Where the Clock Stops
The rule is part of the U.S. aviation regulations (section 25.803 of Title 14 of the Code of Federal Regulations). For an airplane with more than 44 passenger seats, the maker must show that the maximum number of people it can carry, including the crew members the rules require, can get from the airplane to the ground in 90 seconds under simulated emergency conditions [1]. Other countries have their own rules; this article follows the U.S. version, because its test details are public. The European Union Aviation Safety Agency (EASA), which does the same job for Europe, took part in the A380 test described below.
The timing ends when the last occupant has evacuated the airplane and is on the ground (people using a test stand or ramp at an overwing exit count once they are on it, under conditions the appendix sets) [2]. That means the 90 seconds covers more than walking speed. It also includes the time it takes for the evacuation slides to inflate and for people to start moving from their takeoff seats.
The default way to prove this is a full-scale demonstration following a long checklist called Appendix J. The regulator may accept a mix of analysis and smaller tests instead, but only if it finds that this would give data equivalent to a real demonstration [1].
2. Why the Test Uses Only Half the Exits and Keeps the Choice Secret
In a real accident, fire or damage can block some exits. So the test does not let the airplane use all of its doors. Appendix J says the exits used must include one from each pair of exits, must be representative of all the exits on the airplane, and must include at least one at floor level [2]. The text never says "50 percent," but using one of every pair leaves about half of the exits in use.
Nobody is told which doors will be used. No crew member or volunteer may be given a hint beforehand, and the airplane is set up so the active exits are not revealed until the demonstration starts [2]. The crew has to look, decide and guide people on the spot. A simple way to picture the effect is that a design that works only when one favorite door is open would fail this test. That is our own reading, not something the regulation states as its purpose.
3. Who Takes Part in the Test and How the Conditions Make It Harder
The volunteers are meant to be a "representative passenger load of persons in normal health." At least 40 percent must be female, at least 35 percent must be over 50 years old, and at least 15 percent must be women over 50. Three life-size dolls stand in for infants and are not counted in the passenger total. Crew, mechanics and training staff who regularly work on the airplane cannot be used as passengers [2].
The rules then take away helpers. Outside light must be no brighter than 0.3 foot-candles before the emergency lighting comes on, which is about the dimness of a night scene, and the only added light is the airplane's own emergency lighting. About half of the typical carry-on bags, blankets and pillows are spread across aisles and paths to the exits as mild obstacles. Volunteers may not be rehearsed or told the procedures. Nobody who took part in a similar demonstration in the previous six months may join. The flight crew may not physically help people inside the cabin [2].
Put together, the test is designed to look like a crowd of ordinary people meeting the situation for the first time. The regulation does not spell out that goal, so treat that as an interpretation. One thing it clearly does not do is set a share of children or people with disabilities, and that gap matters in chapter 6.
4. The Airbus A380 Test: 873 People Out in 78 Seconds
Could a huge double-deck jet really pass? In March 2006, in Hamburg, Germany, Airbus tried it with the A380, the largest passenger airliner of its time. According to the Flight Safety Foundation's account of the Airbus figures, 873 people (853 passengers and 20 crew) left in 78 seconds, 12 seconds under the limit. They used the eight doors on one side, out of the airplane's 16, and flight attendants from the airline Lufthansa guided the crowd. That works out to roughly 11 people per second, crew included [4].
The evacuation demonstration was certified within days by EASA and by the U.S. Federal Aviation Administration (FAA), and the regulators approved 853 as the passenger maximum [4][5]. News reports of the day rounded the time to about 80 seconds, so 78 should be read as Airbus's own number.
The test was real, and so were the risks. A newspaper report said one person broke a leg and 32 others had minor injuries [5]. That is one reason regulators look for ways to reduce injuries in these tests, and why the rules allow some alternatives.
5. How the Limit Shrank From 120 Seconds to 90
The number was not always 90. A 1965 U.S. rule change (effective March 3, 1965) first required airlines to run full-scale evacuation demonstrations. They used half of the exits, with the others made unusable to imitate blockage, and the time limit was 120 seconds [3].
The FAA's 1995 notice says that amendments effective October 24, 1967, set the limit for the airplane's design approval at 90 seconds and cut the airline operating rule to 90 seconds as well. That notice says the cut came mainly from big improvements in evacuation devices, such as inflatable slides [3]. So the "half of the exits" idea dates back about 60 years, and the 90-second limit followed the equipment: when the tools got better, the standard got tougher.
6. Is 90 Seconds Still a Fair Test on Today's Planes?
Seats have become tighter since the 1960s, and people have asked whether 90 seconds still holds in modern cabins. In 2019 and 2020 the FAA ran evacuation tests on this question. A news report described the volunteers as healthy adults under 60, with no children, older people or people with disabilities, and with nothing on the floor to trip over. U.S. Senator Tammy Duckworth called the conditions completely unrealistic and proposed a bill asking for tests closer to real conditions [6]. Reports of how many people took part and what the results were disagree, so this article does not give those numbers.
What can be said safely is this. The 90-second figure was lowered once because equipment improved, and today people debate whether the test captures modern seating and the full range of passengers. How best to test it is still an open question.
7. What Passengers Can Do to Help the 90 Seconds Work
The number depends on passengers too. The airline association IATA runs a campaign called "Save a Life, Not a Bag," asking people to leave belongings behind in an evacuation, since bags can block aisles and damage slides. In a survey in four countries (the U.S., U.K., United Arab Emirates and Singapore), only 18 percent of passengers knew that airplanes are designed so everyone can evacuate within 90 seconds [7].
You can try a calm version at home. When you board a plane, notice where the nearest exit is. At home, you can walk at a normal pace from your bedroom to the front door and count the seconds. This is only a timing exercise, not an evacuation drill, so do not run, and watch your step. Ask yourself how much of 90 seconds a single hallway uses up.