I checked carry-on rules at 75 airlines. The carry-on wasn't the trap.

I watched someone put their backpack in a sizer at a gate a couple of months ago. It was less than an inch over the personal item rule. The agent apologized, said no, and made her check the bag. She walked away frustrated. I walked away thinking about the data.

That moment kicked off a small project. I pulled the published carry-on and personal item rules for 75 airlines and tested both. The carry-on data came out about how I expected: 14 of 75 airlines reject a standard 22 by 14 by 9 inch suitcase outright on at least one published dimension. The personal item data did not. 24 of 75 airlines do not publish personal item dimensions at all, including Delta, the largest US carrier. Among the 51 that do, the smallest published allowance is roughly one third the volume of the largest.

How I tested

The data lives in a JSON file on my server. 75 airlines, each with a carryOn object and a personalItem object. Both have a published dimensionsIn field for length, width, and height when the airline publishes them, plus a flag for whether basic economy includes the carry-on. The data is verified against airline policy pages between April 14 and April 22, 2026. The full per-airline dataset including everything below is at /data/75-airline-personal-item-trap-2026.csv.

The test bag is 22 by 14 by 9 inches. That is the size most US travelers picture when they hear “carry-on,” and the size most US-made hardshell suitcases ship as. I categorized each airline three ways. Pass means all three published dimensions are at or above 22, 14, and 9. Borderline means a single dimension is within half an inch of the test bag, and the others pass. Fail means at least one dimension is more than half an inch short of the test bag.

I used a dimension-by-dimension test, not the linear-inches sum. Linear inches (length plus width plus height, capped at 45 in most rules) hides the cases where a deep bag fits a tall slot. The dimension test is the version that tracks how a gate sizer actually checks: the bag has to drop into the slot, in any orientation, on all three sides.

The carry-on results

38 airlines pass the test. 23 are borderline. 14 fail outright. The shape is roughly what I expected.

The pass list groups along familiar lines. Three US carriers top the chart with 24-inch length allowances: Southwest, Frontier, and Sun Country. Volaris is the next outlier at 22.4. The big middle band is the 22-inch crowd, mostly US legacy carriers and a mix of European, Asian, and Middle Eastern flag carriers that publish exactly 22 inches in their public-facing rules.

The borderline cluster is more interesting. Almost all of it sits at 21.7 inches. That number is not arbitrary. 21.7 inches is exactly 55 cm. Most non-US carriers set their published rule in centimeters, then convert to inches with a touch of rounding, and 55 cm is the number they use. The standard US carry-on is 22 inches. The standard international carry-on is 55 cm. Those are not the same number. They are 0.3 inches apart. By the published rule, a US-bought 22-inch suitcase is technically out of spec at SWISS, Lufthansa, Air New Zealand, ANA, Japan Airlines, and another fifteen carriers. The gap is small enough that gate sizers usually do not catch it. The gap is also why every traveler who has flown both has felt some version of this confusion.

The outright fails sort along their own logic. Ryanair, Condor, Pegasus, and Vueling all publish 55 by 40 by 20 cm, which is 21.6 by 15.7 by 7.9 inches. The 7.9-inch depth is what kills a 9-inch standard suitcase. Singapore Airlines, Korean Air, ANA, and Japan Airlines publish 9.1-inch depth, just over the bar, with a 21.7 inch length, just under it. The bag does not fit, but only by a fraction on most dimensions. WestJet’s published rule reads 22 by 9 by 14, with width and depth swapped relative to most other carriers; either way, a standard 14-inch-wide bag exceeds the published 9-inch width. Spring Airlines is the strictest in the dataset at 12 inches on the longest dimension, with a 7 kg (15.4 lb) weight cap that would fail a loaded carry-on regardless.

A side fact worth filing away: 18 of 75 airlines block carry-on entirely in basic economy. The list is mostly the European budget set (Ryanair, Pegasus, Vueling, Condor, Norwegian, Eurowings, Wizz Air) plus WestJet and a handful of US ULCCs. On those airlines the suitcase that fits the rule cannot board unless you upgrade the fare. The published rule is gateable. The gate just is not open.

I expected the post to end there

While running the analysis I noticed every airline record had two carry-on fields, not one. The personal item. I expected this to be a copy-paste of the carry-on data, just with smaller numbers. It was not.

Of the 75 airlines, 51 publish personal item dimensions. 24 do not. That includes Delta, the largest US carrier by revenue. Delta’s published rule says only “must fit under the seat in front of you” and lists examples: purses, small backpacks, laptop bags. There is no length, no width, no depth. American, United, Alaska, and most of the big US legacy carriers in this dataset are the same. The personal item rule for the largest airlines in the country is, in writing, “vibes.”

Among the 51 carriers that do publish personal item dimensions, the range is dramatic. IndiGo, India’s largest domestic carrier, allows 13.8 by 9.8 by 5.9 inches, which is roughly the size of a paperback novel laid flat. Volaris in Mexico allows 17.7 by 13.7 by 9.8 inches, roughly a small backpack. The volume difference is about three to one.

The European mainline carriers cluster tightly at 40 by 30 by 10 cm, which converts to 15.7 by 11.8 by 3.9 inches. Air France, KLM, Lufthansa, SAS, SWISS, Austrian, Finnair, Condor, Pegasus, and TAP all sit on or near that line. The 3.9-inch depth is the trap. Most laptop bags are 1.5 to 2 inches deep when empty and 2.5 to 3 inches when packed. A typical small backpack is 5 to 8 inches deep. The published European personal item slot is a slim folio, not a backpack.

The published personal item dimensions are widest at the US ULCCs and a few other carriers: American, Spirit, Frontier, Saudia, and Viva Aerobus all publish 18 by 14 by 8 inches. JetBlue, Breeze, and Sun Country are similar at 17 by 13 by 8 or 9. That is a generous laptop tote, and it is the widest published US standard. Delta and United do not publish dimensions at all but in practice accept similar sizes. The contrast between the European 15.7 by 11.8 by 3.9 rule and the American 18 by 14 by 8 rule is roughly two to one in volume.

If you want to check your specific bag against a specific airline, the data above powers a free checker at vientapps.com/tools/widgets/carry-on-size. It uses the same JSON file and updates on the same cadence.

What this data does not capture

None of the above accounts for gate enforcement. The published rule is the floor, not the ceiling. A bag that meets every published number can still get rejected at the gate. A bag that fails by half an inch can fly fine for years.

Ryanair will measure your bag and weigh it before you board. They have to: the fees on non-compliant bags are how the airline makes money. Lufthansa rarely measures and almost never weighs. American gate agents will eyeball most bags and only test in the sizer if the bag clearly looks oversized or the flight is full. The same airline can be strict at one station and loose at another. Summer routes to Europe are stricter than winter routes to anywhere. The flight that ramps up after a missed connection is stricter than the early morning bank. Whether your bag flies often depends on which side of someone’s bad day you arrive on.

The dataset is the scoreboard. The actual game is played at the jet bridge by a person whose mood you cannot predict. The gate moment I started this post with, where the bag was less than an inch off and the agent said no, is the part the data cannot reach.

Why I built the dataset

I built this because I needed structured airline data for the comparison pages on my travel site. There are 56 airline-vs-airline pages that all needed bag fees, carry-on dimensions, basic-economy rules, and a few other fields. Pulling those numbers from prose every time was a non-starter. Once it was structured, exposing it as a public tool was the obvious next step. The widget powering the bag size checker is the same JSON file behind this post and behind the comparison pages. There is one source of truth, and it gets refreshed on a 30-day cadence against the airlines’ own published policy pages.

That is also why this post can be specific. Every number above traces to a record in the file, and every record has a lastVerified date and a sourceUrl. The CSV linked at the top has all of that.

What I am still figuring out

The unanswered question is whether the airlines that do not publish personal item dimensions are stricter or looser at the gate than the ones that do. The Delta-style “fits under the seat” rule is permissive on paper and may translate to permissive in practice. Or it may not. I do not have that data. The next analysis I want to run is a sample of routes per airline at the gate, with a real bag, watching what gets through and what does not. That is a longer project and probably needs help from people who fly more than I do.

If you have a story about a personal item that got rejected (or one that should have been rejected and breezed through), I want to hear it. The stories are the part of this dataset that is missing.