Flight booking myths

For US travelers

What actually makes a flight cheaper, and what's just folklore.

Book on a Tuesday. Wait exactly 21 days. Clear your cookies before you buy. Every one of these gets repeated as settled fact, and most of them do not hold up the way the internet says they do.

This page checks six of the loudest booking myths against the actual studies, the actual regulation, and the actual law, and says plainly where the evidence runs out. Some of it is a real, small pattern. Some of it is a genuine legal protection with real conditions attached. And at least one popular claim is a misreading of a live federal action that does not do what people are currently saying it does.

What this page will and won't tell you

This is not a replacement set of booking instructions. Where the evidence supports a real pattern, we say so, and how strong it is. Where the field disagrees with itself, we say that too, instead of picking a winner to sound decisive. Nowhere below will you find a "book on this day" instruction, a VPN playbook, or a claim that a federal agency has started policing airline prices. Two of those would be dishonest. The third would just be wrong.

Part of the flights set: what your airline owes youRefund rights when the airline is the one that cancels or changes your flight, not a booking strategy.Part of the flights set: what your fare doesn't includeBags, seats, basic economy, and the self-transfer risk this page points to in section 5.

1. The best day to book, and the best day to fly, are not the same question

These get conflated constantly, and they are two separate actions with two separate effects. Booking on a particular day of the week shifts a fare by a percent or two, if it moves it at all. Flying on a particular day of the week moves it by a lot more. Neither one is a myth exactly, but only one of them is worth planning around.

Three studies, three different "best" days, in the last twelve months

Google's own analysis, published September 2025, found Tuesday bookings averaged about 1.3% cheaper than Sunday. Expedia's 2026 Air Hacks report, published five months later in February 2026, found Friday was the cheapest day to book, by about 3% over Sunday, and explicitly moved off the day it had favored the year before. KAYAK's most recent data points to a different day again. These are not three measurements converging on one answer. They are three different booking pools, over three different windows, and the "winning" day has changed at least once a year for as long as we've checked. Whatever a single site tells you the cheapest day to book is this month, do not expect it to hold next year, or to say anything about your specific route.

The one piece that does hold up: when you fly matters more than when you book

Every study above found a bigger, more consistent effect from the day you actually travel. Google put midweek departures at roughly 13% cheaper than weekend ones. Expedia's own 2026 data found flying on a Friday, rather than the day it now says is cheapest to book on, saved travelers noticeably more than the booking-day effect did. If you're only going to optimize one of the two, it's the day you fly, not the day you buy.

Source: Google's 2025 travel trends, Expedia's 2026 Air Hacks report, and KAYAK's current data.

2. There's a shape to the advance-booking window. There's no magic number.

Set that reminder for "book 21 days out" and you're treating one provider's average as a rule for your trip. Every major study below agrees on the shape of the curve: not the day a fare first opens, and not the final week before departure, but a mid-range window somewhere in between. None of them agree on the exact number of days, because none of them are measuring the same routes over the same period.

Read across those four and the honest takeaway is "somewhere in the three-to-five-week range for a domestic trip," not any single day count. Treat whichever number a headline gives you as one provider's average, not a countdown to set a calendar reminder around.

A number we're deliberately not printing. Expedia's own published international-savings figure does not reconcile with the underlying numbers in its own report; the appendix implies two different totals depending on whether you read the stated figure or do the subtraction yourself. Rather than pick one and present it as verified, we're leaving that specific number off this page until it's corrected at the source.

None of this means last-minute fares are always expensive; they're rarer than they used to be, not extinct. A seat that would otherwise fly empty can still get discounted close to departure, especially off-peak. That's a real exception to the general shape above, not a strategy to plan a trip around.

Source: the same Google and Expedia reports above, plus KAYAK's and CheapAir's own advance-purchase data.

3. Cookies, incognito, and VPNs: what's actually been tested, and what's genuinely still open

This section is really two separate claims wearing one hoodie. The first is "clearing your cookies gets you a cheaper fare," which has been tested directly and doesn't hold up. The second is "airlines never use anything about you to set a price," which nobody can honestly say either, because AI-driven pricing is real, growing, and not yet fully understood from the outside. We're not going to soften either half to make a cleaner story.

What's been tested: clearing cookies

The most-cited test of this, and it's from 2016

The test people are usually pointing to, whether they know it or not, is a 2016 Consumer Reports investigation: 372 paired flight searches across 9 airline and travel sites, one browser with cookies intact, one freshly scrubbed. The two versions came back with identical prices the large majority of the time. Where they differed, the scrubbed browser wasn't reliably cheaper; it came back more expensive almost as often as it came back cheaper. That's a decade-old test of one specific mechanism, not a live 2026 audit, and it predates most of the AI pricing work below by nearly a decade. Treat it as historical evidence against the cookie-clearing folklore specifically, not as a verdict on everything airlines might do with data today.

What's genuinely real, and gets confused with cookie-clearing: being shown a different price based on your apparent currency or location, which runs through your card network and merchant country handling, not a tracking cookie, and can be overridden the other way by how your own card processes a foreign charge. None of this is a repeatable hack. It's closer to noise than to a lever you can pull on purpose.

Source: Consumer Reports, 2016.

What's genuinely unsettled: AI pricing and your personal data

We won't tell you airlines never use anything about you to set a price. AI-driven fare pricing in this industry is not speculative; it's being deployed, named, and dated.

What these systems are confirmed to do is set prices dynamically, off market demand, route, timing, and inventory. What's actually disputed, as of late 2026, is whether any of them price you differently from the traveler next to you based on who you are, rather than what's left to sell. That's genuinely unproven in either direction, not a settled "yes" and not a settled "no."

Source: Delta's own statement, PROS and Lufthansa Group, and Amadeus and SAS.

Get the Washington scrutiny right, or don't publish it. Two separate things are happening right now, and conflating them is the easiest way to get this topic wrong. First: Rep. Frank Pallone, ranking member of the House Energy and Commerce Committee, sent letters to eight major US airlines in August 2026 asking how they use personal data in pricing, alongside a pending JetBlue class action alleging "surveillance pricing." Both are allegations under active inquiry, not findings; no airline has been shown, in any governmental or judicial finding, to price individual travelers off their personal data. Second, and separate from that: the FTC's own personalized-pricing action, proposed August 19, 2026 with comments due September 25, 2026, is a proposed enforcement policy statement, not a final rule, and it's aimed generally at retail and grocery pricing practices. It does not newly regulate airlines. Under a long-standing carve-out in Section 5(a)(2) of the FTC Act, air carriers sit outside the FTC's Section 5 jurisdiction entirely, an exemption that predates this proposal by decades. Reporting the FTC action as a new airline crackdown is the specific mistake to avoid here.

Source: Rep. Pallone's letters, the FTC's proposed statement and its comment extension, and legal analysis of the airline carve-out.

This whole section is dated deliberately, as of late 2026. It's the fastest-moving part of this page by far: check the byline date below before treating anything above, the deployments or the regulatory posture, as still current.

4. A mistake fare isn't automatically yours to keep

A wildly underpriced "error fare" doesn't come with a guarantee the airline has to honor it. Under DOT's 2015 Mistaken Fare Policy Statement, still the operative guidance with no rule since replacing it, the Department's enforcement office won't act against a carrier that cancels a genuine mistaken fare, as long as it refunds the ticket price and reimburses reasonable, verifiable, out-of-pocket costs a traveler incurred relying on the booking, a hotel deposit or a tour reservation, for example, not the value of the trip itself.

That's a real, still-standing protection, and it comes with real conditions: the fare has to actually be a demonstrable mistake, and your losses have to be documented. It isn't a blank check for an airline to void any fare it later regrets, and it isn't a strategy this page is going to teach you to go hunting for.

Source: DOT's 2015 Mistaken Fare Policy Statement.

5. Two one-way tickets, or a connection someone else stitched together

"Two one-ways always beat a round trip" is wrong exactly as often as it's right. Sometimes splitting a trip into two separate one-way fares comes in cheaper. Sometimes the round-trip fare undercuts it. There's no rule that reliably favors one over the other, so the only real answer is to price both for your actual dates before you book either.

A separate, more consequential question: whether the connection in your itinerary is one ticket or two. Some of the cheapest-looking connections a search site shows you are actually two separate tickets on two airlines with no agreement between them, stitched together by the search engine itself, not sold that way by either carrier. Book that way and you give up through-ticket protection: if your first flight runs late, the second airline owes you nothing, because as far as its system knows, you simply didn't show up. That risk isn't automatic doom, since plenty of self-transfer connections work out fine, but it isn't automatically protected either, and knowing which kind of itinerary you're holding before you fly is the whole game.

The fuller version of the self-transfer risk lives on the fee-mechanics pageSame fact, one home: what a stitched-together itinerary actually looks like and why it isn't automatically protected.

6. And the 24-hour rule doesn't "lock" a fare either

One more piece of booking folklore worth a single line here: the DOT's 24-hour rule is a hold-or-refund right, not a price lock, the airline picks which of the two it offers, and it only reaches bookings made directly with the airline. We cover the mechanics, and where it doesn't reach, in full on the rights page rather than repeat it here.

The full 24-hour rule lives on the rights pageWhat it actually guarantees, which bookings it reaches, and the difference between a hold and a refund.

Written and kept current by the True Trip Costs desk. The personalized-pricing facts on this page are the fastest-rotting on the site and are re-checked quarterly at minimum, and immediately on FTC, DOT, or deployment news; the day-of-week and advance-booking study figures are re-checked as each provider publishes its next annual report. Who writes these, and how they are checked →

Sources, and how this page was checked

Last checkedSep 2026

Primary sources

Our own readThe FTC's proposed statement does not regulate airlines: air carriers sit outside FTC Section 5 jurisdiction under a decades-old carve-out, separate from anything in this specific proposal, and reporting on the topic has sometimes blurred that line. The Pallone inquiry and the JetBlue litigation are both allegations under active process, not adjudicated findings, and we've stated them that way. Delta's own description of its AI pricing work explicitly denies individualized targeting, which we've included rather than let the "AI pricing exists" fact imply more than the evidence supports. Expedia's own published international advance-booking savings figure doesn't reconcile against its stated total, so it's omitted rather than repeated as verified. The self-transfer risk in section 5 and the 24-hour rule in section 6 both already have a fuller home on this site's other flight pages; this page states the fact once and links out rather than keeping two versions that could drift apart.

Check the evidence, not the folklore

None of these six myths resolve into a clean hack, and that's the honest answer, not a dodge. Where a real pattern exists, it's small and it moves every year. Where a real legal protection exists, it comes with conditions worth knowing. And where the internet has started repeating a regulatory story that isn't quite true, correcting it plainly is worth more than a tidy headline. For the money side of the trip once you've actually booked, see what your airline owes you when it changes your plans, and what your fare didn't include in the first place.