Short answer. The quicker orders in this comparison spread passengers along the cabin and seat window passengers before aisle passengers. The Steffen method does both and was derived by optimisation for a single-aisle aircraft boarded through one door. In this simulator, it takes less than half the time of back-to-front zones with the same passengers. WILMA (window, middle, aisle) is the simpler runner-up, and a plain random queue beats the zone systems tested here.
The ranking
Simulator result: Reference flight: easyJet Airbus A320, 175 passengers (about 94% full), typical cabin luggage, forward door only. Each time is the median of eight seeded runs (seeds 20260801 to 20260808); every method sees exactly the same passengers and bags.
| Method | How it works | Boarding time |
|---|---|---|
| Steffen method | Alternate rows, one side at a time, windows first | 8:21 |
| WILMA | All windows, then all middles, then all aisles | 12:31 |
| Reverse pyramid | Rear windows first, front aisles last, in diagonal bands | 13:10 |
| Random queue | No groups, one line in random order | 14:50 |
| Rotating block | Rear block, then front block, then the next rear block | 18:45 |
| Back-to-front | Rear zone first, then forward, zone by zone | 19:15 |
| Airline groups | Assistance, premium rows, then rear to front | 19:26 |
| Front-to-back | Front zone first | 23:15 |
| Open seating | No assigned seats, take the best one left | 25:29 |
Simulator result: Measured this way, the Steffen method took about 57% less time than back-to-front and about 44% less than a random queue. WILMA took about 16% less than random. The Cincin method, the simulator's experimental multi-door extension of Steffen, is identical to Steffen when there is only one door and is left out of this table for that reason.
Why the fast methods are fast
Boarding slows down largely because passengers stop, rather than because they walk slowly. Two common causes are waiting behind someone lifting a bag into the bin (aisle interference) and waiting while someone already seated stands up to let you in (seat interference). Why does boarding take so long? explains both.
Spreading people out reduces the first. If passengers in the aisle are heading for rows far apart, they can stow their bags at the same time instead of queueing behind one another. The Steffen method generally places consecutive passengers two rows apart, allowing a wave of passengers to stow bags in parallel.
Windows first reduces the second. Seating window passengers before aisle passengers means fewer people need to pass a seated neighbour. Simulator result: On the reference flight, WILMA and Steffen each had a median of 3 squeeze-bys, against 64 for back-to-front and 66 for a random queue.
The zone systems offer less of either benefit. Back-to-front sends everyone in a zone to the same few rows, where they queue behind one another while much of the aisle remains empty.
Research finding: Steffen's ordering was derived by optimisation and later tested in a physical experiment with volunteers on a 12-row, single-aisle mock-up boarded through one forward door. That experiment supported the broad ranking of parallelised methods over block boarding. It does not validate the absolute times above, which are outputs of this model.
Why random beats back-to-front
It sounds counterintuitive, but a random queue spreads passengers along the cabin by accident. Some are heading for row 3, some for row 28, so some can stow bags in parallel. Back-to-front removes that accidental spread. Research finding: Published boarding studies also report that block boarding is not clearly better than random boarding.
Does the fastest method change with the aircraft?
Simulator result: On a completely full Ryanair 737-800 (189 passengers, typical luggage, forward door, seeds 20260801 to 20260808), the same four methods led in the same order: Steffen 8:57, WILMA 13:09, reverse pyramid 13:50 and random 15:47. Further down, back-to-front and airline groups swapped places. The ranking changes more when two doors are used, because a sequence designed from the nose sends the rear-door stream the wrong way. Simulator result: On a United 787-9 boarded through front and rear doors (231 passengers), the Steffen method took 7:40 and the door-aware Cincin method 4:48. Those multi-door results are outputs of this model only; the published experiments did not cover them.
Frequently asked questions
What is the fastest boarding method?
Simulator result: For a single-aisle aircraft with one door, the Steffen method is the fastest of the implemented strategies (Cincin ties with it, because the two are identical with one door). That is consistent with the published optimisation work. It is not in common use at gates because it is designed around passengers lining up in an exact order; see Why don't airlines use the fastest boarding method?.
Is WILMA faster than back-to-front?
Simulator result: Yes, on the reference flight WILMA took 12:31 and back-to-front 19:15, about 35% less. It removes almost all seat interference and is easy to announce, which is why versions of it are used for economy groups by some airlines.
Why is open seating the slowest in the simulator?
Model assumption: In this model, each open-seating passenger takes the best free seat near the door when they step aboard, so early boarders fill the front and later ones have to walk past them. Real open-seating operations often add their own discipline, such as numbered lineups, which the model does not include.
Updated 2026-09-25. Figures labelled simulator result are outputs of this site's model at the settings stated, not airline measurements. What the model leaves out.