Twenty carry-ons against six budget-airline sizers

Bag 08 is a hard-shell spinner that its maker sells as a carry-on. With wheels it stands 575 mm tall, 405 mm wide and 252 mm deep once I had put 7 kg in it. Of the six frames I built, the biggest is 560 x 450 x 250 mm. Bag 08 missed that one by 15 mm on length and 2 mm on depth, and it missed the other five by more. In three days of measuring, it did not fit anything.

That is an extreme case, and the bag is not a bad one. It rolls well and the zips are good. It just does not fit the boxes that a budget airline sets at the gate, and the product page gives no hint of it. So I took 20 carry-ons, packed them the same way and measured how many of them go through six frames cut to budget-airline sizes.

I used six open-fronted frames. Each is built to a limit that one budget airline publishes for a cabin bag or a free under-seat bag, rounded to the nearest 10 mm. I call them Airline A to F and I do not name the airlines, for two reasons: their rules change, and my frames are not their frames. Real gate sizers are looser or tighter in ways I cannot reproduce, such as a sloping back or a bar that flexes.

In millimeters, length by width by depth:

• Airline A: 400 x 300 x 200

• Airline B: 450 x 360 x 200

• Airline C: 500 x 400 x 200

• Airline D: 550 x 350 x 250

• Airline E: 550 x 400 x 200

• Airline F: 560 x 450 x 250

Frames A and E are the two steel frames, 40 x 30 x 20 cm and 55 x 40 x 20 cm inside. The carry-on guide for budget airlines explains how to check your own bag against sizes like these. I built the other four from 18 mm plywood for this test. D and E are not nested: D is 50 mm deeper and 50 mm narrower. That is what makes airline sizes awkward, because a bag can pass one and fail another that looks bigger. I measured each frame at every corner and it held within 2 mm.

The 20 bags were bought at retail between January and July 2026, one unit each, and every one is sold by its maker as a carry-on. I measured them in the Manchester workshop on Monday 10 to Wednesday 12 August 2026. There are 8 hard-shell spinners, 6 soft-sided wheeled cases, 3 soft duffles without wheels and 3 backpacks, numbered Bag 01 to Bag 20 in that order. I use codes instead of names so that the method gets read.

Standard fill is 7 kg in every bag: a week of clothes in two cubes, a 14-inch laptop, a wash bag, a pair of trainers and a rain jacket. Zips closed, any expander closed, telescopic handle collapsed. Three of the hard shells had expanders, and I kept them shut. Full fill is for the 12 soft bags only, because a hard shell cannot be stuffed. I added clothes until the zip would close with one firm press, without sitting on the lid. That took them to 9.4 to 11.6 kg.

For each bag I measured length from the floor to the highest point, wheels and top handle housing included, then width at the widest point and depth at the thickest point, with zips and front pockets included. I used a steel rule and a square against a flat bench, and for soft bags a flat board lowered on top with a 1 kg mass on it. I measured twice, and the two readings never differed by more than 3 mm. A bag passed a frame if it went in, in any orientation, with one firm push of one hand, and came out again. The tape and the frames agreed on every bag at standard fill and every soft bag at full fill.

Samuel Adeyemi, who reviews our data stories, re-ran every pass mark from my raw table and found no differences. The full protocol is on our how we check page.

At standard fill, 17 of 20 bags passed at least one frame. The three that passed none were Bag 01, Bag 04 and Bag 08, all hard shells. Only one bag passed all six: Bag 19, a small backpack that measures 400 x 290 x 190 mm. Stuffed to the zip, it still passed D and F, but failed A, B, C and E.

Counts at full fill include the 8 hard shells at their one fill, so only the soft bags change. The pattern is plain. The three smallest frames, A to C, passed 9 bag and frame pairs between them, and the largest, F, passed 17. Frame E, the 55 x 40 x 20 cm size that many people think of as the standard cabin bag, passed 7.

Depth sorted the bags more than any other number. Thirteen of the 20 were deeper than 200 mm at standard fill, and the four frames that are 200 mm deep (A, B, C and E) let none of those 13 through. Every bag that passed E was one of the 7 that measure 200 mm or less. The 8 hard shells run from 198 to 252 mm deep, with a median of 230. Only Bag 05, at 198 mm, got under 200, and it passed E and F. Bag 07 missed C by 8 mm on depth, and Bag 10 missed E by 5 mm. Bag 15 missed E by 10.

D and E split three bags. D passed Bag 07 and Bag 15, which E refused on depth, and E passed Bag 05, which D refused on width by 10 mm. A traveler who tested a bag against the wrong airline's size would have been right or wrong by luck.

The 12 soft bags look better at 7 kg than the hard shells, and they are. At standard fill they had 34 frame passes between them. Stuffed until the zip closed, they had 13, a loss of 21. Depth grew by a median of 41 mm, from 25 mm on the least-changed bag to 50 mm on the worst. Width grew by a median of 15 mm. The unwheeled bags also grew 20 mm in length, because stuffing a duffle pulls its ends outward.

Four soft bags that passed F at 7 kg passed nothing at all when stuffed: Bag 09, Bag 11, Bag 13 and Bag 17. That is the fair price of soft: a bag that squashes into a sizer can also bulge past it. The people who get caught are the ones who pack to the zip because the bag feels roomy. At standard fill the soft wheeled cases passed F every time, 6 of 6. Stuffed, 3 of 6 did.

If you take one thing from this table, it is that a soft bag has a fill level, and the fill level is the difference between a pass and a fee. Pack it to 7 kg and see how it measures. A bag that passes at 7 kg and fails at 10 kg is telling you where to stop.

The 14 wheeled bags, the hard shells and the soft cases, are where the argument starts. A maker's listed length often leaves out the wheels and the handle housing. Nine of the 14 product pages I read gave a length without them, and in at least two cases the only mention was a footnote. Some airlines, easyJet among them, say wheels and handles are included, so I measured to the highest and lowest point of the bag.

Wheels and handles added a median 60 mm to the body length, in a range of 50 to 65 mm. That is enough to move bags across a line. If I had measured the body only, 13 of the 14 wheeled bags would have passed at least one frame, and with wheels and handles it was 11. Bag 01 and Bag 04 would have passed F. Bag 12 and Bag 14 would have passed B, and Bag 05 would have passed C. None of that survives a gate agent who looks at the wheels.

There is a trade here that I felt on the cobbles of Lisbon and Porto in July, when I rolled Bags 05, 07, 12 and 14 over them for five days. The two bags that added the least, Bag 12 and Bag 14 at 50 mm each, have small recessed wheels. They also caught on the joints between the stones more often and were louder than the larger-wheeled hard shells. Two bags is an anecdote, not a finding. But it matches what the numbers suggest: a bag that adds little length has less wheel to work with.

The handle is the quieter argument. On Bag 13 the telescopic handle housing added 9 mm to the depth at the back. That was invisible with the handle collapsed unless you laid a rule across it, and it is the kind of thing a sizer finds. I counted it, because the frame would have.

These are my reconstructions of six sets of published limits, not the real sizers. Some gates measure, many do not, and an agent has discretion in both directions. I tested one unit of each bag, so I cannot say what a different batch measures, and a soft bag settles after a few trips. Twenty bags is a small sample that leans toward bags sold as carry-ons by well-stocked makers. A hard-shell spinner from a maker I did not buy may be shallower than all of mine.

I did not weigh these bags against any airline's weight limit, and many airlines have one. The 7 kg standard load is a convenient number, not a rule. For the weight side of the story, see our backpack weighing. The raw table of measurements and pass marks is available on request.

The sentence I would quote: at 7 kg only 1 of 20 carry-ons passed all six frames and 17 passed at least one, and stuffing soft bags cut that to 13. Which of these applies to you depends on your airline's number.

If your airline's frame is 200 mm deep, like A, B, C and E, buy a soft backpack or a duffle and measure it packed. Of the 8 hard shells, only one came in at 200 mm or under. Do not rely on a hard shell for these airlines unless the listed depth, with the pockets, is at most 200 mm.

If your frame is 250 mm deep, like D and F, a hard shell or a soft wheeled case can work. Five of the 8 hard shells passed F and all 6 soft wheeled cases did at 7 kg. Check the length with wheels, because Bag 01 missed F by 5 mm on exactly that.

If you do not know the size, read the limit on your booking and measure your own bag, loaded, with wheels, handle and every pocket, to the nearest centimeter. Add 4 cm of depth if it is soft and you tend to fill it. A tape measure at home will tell you more than any product page.