Should You Book Now or Wait? When to Book Flights and Hotels

Everyone asks this question, usually around eleven at night, with four tabs open and a price that was €40 cheaper yesterday. And almost every answer you find online is folklore dressed as data: book on a Tuesday, clear your cookies, never book on a weekend, wait until six weeks out.

We went through the actual numbers. Expedia's booking data, Google Flights' four-year price history, a hotel rate study, a European hotel distribution report, and the consumer regulations that decide whether you can change your mind. Some of it is genuinely useful. Some of it flatly contradicts itself, which is the most useful finding of all.

Should You Book Now or Wait? The Short Answer

The five things behind that answer:

• There is no magic day of the week to book. Shopping on a Tuesday is about 1.3% cheaper than the worst day, according to Google Flights' own analysis of four years of data. That is noise. Ignore it.
• Flights and hotels move for opposite reasons. A flight generally gets more expensive as seats sell, because the cheap inventory is finite. A hotel room gets cheaper when the night is not selling, because an empty room earns nothing. Same trip, two different strategies.
• The reliable play is not timing. It is optionality. Book something refundable early enough to get what you actually want, then re-check it twice on a schedule. You capture the drops without gambling on them.
• Waiting reliably loses when supply is fixed and demand is not. Events, school holidays, small properties, and any trip where you need a room that sleeps four.
• Set one alarm: 48 hours before your free-cancellation deadline. That single calendar entry is worth more than every "best time to book" study combined.

That is the whole article in five bullets. The rest is why, and the edge cases where it breaks.

Three Myths About the Best Time to Book

Is There a Cheapest Day to Book a Flight?

No. The gap between the best and worst day to shop is somewhere between 1% and 3%, and the two largest studies do not agree on which day wins.

The Tuesday myth has survived a decade past its expiry date. It comes from an era when airlines loaded fare changes into distribution systems in weekly batches, and a sale filed on Monday night showed up across the market on Tuesday. Fares now reprice continuously, many times a day.

Google Flights analysed four years of its own data and found that shopping on a Tuesday came out about 1.3% cheaper than Sunday, the most expensive day to shop. Expedia's 2026 report, working from a different dataset, landed on Friday as the cheapest day to book, 3% cheaper than Sunday.

When two large datasets disagree about the direction of an effect that small, the effect is not real at the level of your individual trip. Book on the day you decide. The day of the week is not the variable.

Does Incognito Mode Make Flights Cheaper?

No. Private browsing does not change the fare you are shown, and neither does clearing your cookies.

Consumer Reports ran 372 flight searches and found private browsing produced a cheaper fare 7% of the time and a more expensive one 5% of the time. The other 88% were identical. That is the signature of random noise, not of a site punishing you for looking.

What people experience is real, though, which is why the myth is so durable. The cheap fare bucket sold out between your two searches. Or the currency or point of sale changed. Or you are seeing a cached price on the first search and a live one on the second. The price moved. It just did not move because of you.

Do Prices Always Go Up Closer to Departure?

No. Flight prices usually rise as seats sell, but hotel prices frequently fall in the last two weeks, and airlines themselves sometimes reopen cheaper fares.

A hotel room is the most perishable product in travel: tonight's empty room cannot be sold tomorrow, so a property sitting at 60% occupancy three days out has a strong incentive to drop the rate. A NerdWallet analysis of more than 2,500 hotel rates collected between 2019 and 2021 found rooms were around 13% cheaper booked 15 days ahead than four months ahead, roughly $30 a night, and around 22% cheaper for luxury properties.

Flights break the rule too, in a specific way we get to next.

What Makes Flight Prices Go Up or Down?

Flight prices are set by how fast a specific flight is filling compared with the airline's forecast for it. Seats are sold in fixed-size price tiers. When a tier sells out, the price steps up. When a flight is selling badly, the airline can reopen a cheaper tier and the price steps back down.

Seats Are Sold in Buckets, and Buckets Are a Staircase

An airline does not put 180 seats on sale at one price. It divides them into fare classes, sometimes called buckets, each with its own price and its own allocation. The cheapest bucket has the fewest seats. When it sells out, it closes and the next one opens, at a higher price.

This is why airfare moves in steps rather than a smooth curve, and why it can jump €80 overnight with nothing else changing in the world. Six people bought the last of a bucket while you slept.

It is also why the price can go down. Revenue management systems watch how a flight is filling against its historical pace. A flight selling behind pace gets its cheaper buckets reopened to stimulate demand. A flight selling ahead of pace has them closed early. The airline is not tracking you. It is tracking the flight.

Four Other Things That Move a Fare

Competition on the route. The single biggest determinant of what a route costs. Two carriers flying a city pair daily produce very different pricing behaviour than a monopoly route with one flight a day. When you see an implausibly good fare, it is usually a carrier defending market share, not a glitch.

Frequency. More departures a day means more chances the algorithm is trying to fill a specific one. A route with six daily flights has soft spots. A route with one does not.

Season, and your departure day. Expedia's data has August domestic US travel running about 29% cheaper than December, roughly $120 a ticket, and January about 19% cheaper. Departure day matters more than booking day: Google Flights found flying Monday to Wednesday came in about 13% cheaper than Friday to Sunday, and about 20% cheaper on domestic routes. Expedia put Friday departures 8% below Sunday overall, and Tuesday domestic departures 14% below Sunday. Move your departure by a day. It beats every booking-timing trick in this article.

Connections. Google Flights found itineraries with a connection ran roughly 25% cheaper than nonstop. Whether that is a saving depends entirely on the connection, which is a different article and a real cost we do not think should be waved away.

What Makes Hotel Prices Go Up or Down?

Hotel prices are driven almost entirely by occupancy: how full one specific night is right now, compared with where the hotel expected to be. Everything else is a modifier on that single number.

Hotels feel more mysterious than flights and are actually simpler. A revenue manager is looking at a screen of dates. Each date has a booked position and a pace, meaning how it compares to the same point in the cycle last year. Dates ahead of pace get pushed up. Dates behind pace get pushed down, usually late, usually in the last two weeks.

The modifiers:

• Events. A conference, a festival, a marathon, a religious holiday. Citywide compression is the fastest and most brutal price move in travel. A €140 room becomes a €420 room, and no strategy in this article will save you.
• The weekday inversion. Business hotels are expensive Monday to Thursday and cheap on Saturday. Leisure hotels are the reverse. In cities that serve both, the same star rating can differ by 60% between two adjacent hotels on the same night, purely because of who each one was built for.
• Cancellation churn. This one is underrated. A study by D-EDGE covering 680 European hotels found cancellation rates rose from 32.9% in 2014 to 39.6% in 2018, and reached 50% on Booking Holdings channels. Direct bookings cancelled at 18.2%. Roughly two in five reservations on some channels evaporate, which means rooms come back into inventory continuously, often at the current rate rather than the old one.

That last point is the mechanism behind the entire "check again later" strategy. Inventory is not a fixed pool draining toward zero. It refills.

Why "Book 15 Days Out" Is a Trap, Even Though It Is True On Average

The 15-day figure is an average across soft dates and compressed dates, and those two behave in opposite ways. On a soft date, waiting works beautifully. On a compressed date, waiting means expensive or sold out.

Booking 15 days out is a strategy that wins small and often and loses large and occasionally, which is exactly the risk profile you do not want on the one week a year you take off work.

The lead-time data supports treating this as a range, not a rule. D-EDGE put the average booking lead time across channels at 39 days, with direct bookings at 41 and Booking.com at 35. That is the middle of a very wide distribution, not a target.

How Far in Advance Should You Book? The Two Biggest Datasets Disagree

There is no single correct booking window. The two largest published analyses contradict each other by about eighty days on international trips, and understanding why is more useful than either number.

Expedia, using its own booking data from December 2024 to November 2025, reports that international flights booked 8 to 15 days ahead saved an average of $225 versus booking six months out, with a second good window at 31 to 45 days.

Google Flights, using four years of price history, reports that fares to Europe are lowest around 72 to 94 days out, and that on US-to-Europe routes there is no longer a dip closer to departure at all.

These are not slightly different. One says book two weeks out, the other says book three months out. Both are large, competent, well-funded analyses.

They disagree because they measure different things, and this is the single most useful idea in this article:

• Expedia is reporting what people paid. That is a mix. It includes routes, cabins, seasons and traveller types all averaged together. Late bookings skew toward flexible leisure travellers chasing discounted inventory on soft routes. The average says "people who booked late paid less," which is not the same as "you will pay less if you wait."
• Google is reporting what one itinerary cost over time. That is a price history for a given trip. It answers the question you are actually asking.

What to take from it: treat any published booking window as weak background information about a market, not a prediction about your trip. The only price history that matters is the one for your dates, your route, and your hotel. Which is why the strategy below is built on watching your own trip rather than obeying somebody's average.

When Is the Best Time to Book a Hotel? Book Refundable, Then Re-Check Twice

Book a refundable room as soon as you decide on the trip, then check the rate twice: around 45 days out and around 14 days out. Rebook only if the drop is bigger than what the flexible rate cost you. This works whichever way the price moves, which is the point, because nobody can tell you which way it will move.

Step 1. Book Refundable, Early Enough to Get What You Want

Book when you have decided on the trip, not when a study says to. The point of booking early is not the price. It is that the good hotel in the right neighbourhood with a room that fits your family is a scarcer thing than a cheap rate, and it goes first.

Step 2. Put Two Re-Checks in the Calendar. Two, Not Fourteen.

Two checks. Setting a daily price alert on your own trip is how you turn a €60 saving into three weeks of low-grade anxiety, and it does not improve the outcome.

Step 3. Rebook Only If the Drop Beats the Premium

Rebook only if the drop is larger than the refundable premium you paid, and only if the new booking is genuinely the same thing.

That second clause catches people. The cheaper rate is frequently a different room type, a lower floor, a non-refundable rate, breakfast removed, or a third-party rate where the hotel does not hold your reservation directly. Compare the line items, not the headline number.

This is, roughly, the arithmetic we run for people who would rather not run it themselves — it is what the shortlists on how Escapada works are doing when they say what a room actually costs once the conditions are counted.

Should You Book a Refundable Hotel Rate?

Take the flexible rate when your plans are uncertain, the date looks soft, or the premium is under about 10%. Take the non-refundable rate when the date is obviously compressed, you are certain you are going, and the discount is 20% or more.

Here is the honest maths. The flexible rate is an option contract, priced by someone with better data than you have. If flexible costs 12% more and rates on your kind of date drop 13% on average, you are close to break-even before you have spent any time on it. On a sold-out date you were never going to rebook lower, so a flexible rate there is an option you will never exercise.

The Best Time to Book a Flight Is Different

For flights: use the 24-hour rule below as a genuine free hold, decide inside that window, then book and stop looking. If your route has real competition and daily frequency, one check at 45 days is enough. Beyond that, watching the price is a hobby, not a strategy.

Can You Cancel and Rebook? The Rules That Make This Work

In the US you get 24 hours to cancel a flight bought directly from an airline. In the EU there is no equivalent right, and your only protection is the fare rule or hotel rate you chose. That difference decides how much of the strategy above is available to you.

Two practical notes on hotels. First, "free cancellation" on some third-party rates means the intermediary refunds you, occasionally minus a service fee, rather than the hotel releasing the booking, and those are exactly the rates that look cheapest in a results list. Second, the deadline is a timestamp, not a day. 6pm local time on the 12th means 6pm on the 12th where the hotel is.

When Waiting to Book Reliably Costs You More

Book immediately, and skip the re-checks, in these five situations.

1. Fixed-date, fixed-place events. A wedding, a conference, a marathon, a festival, New Year in a small town. Demand is inelastic and supply is a hard number. Prices only go one direction.
2. Peak school holidays on the routes that serve them. Every family in a country is bidding for the same two weeks on the same handful of flights.
3. Small inventory. An island with three hotels. A lodge with eight rooms. A sleeper train. A single daily flight. Bucket logic gets vicious fast when the buckets are tiny.
4. You need a room that fits four. In most of Europe, rooms that legally sleep four are a small fraction of a property's inventory. You are not competing on price, you are competing for the eleven rooms in the city that fit you. This is the most underrated one on the list.
5. Inside 21 days on a thin international route. The cheap buckets are gone and there is no competitor to force them back open.

When It Is Safe to Wait Before Booking

Waiting carries little risk in these four situations.

1. Big-city, high-supply destinations in shoulder season. Lisbon in November, Berlin in February. There is more room than demand and the rate will soften.
2. Routes with several carriers and daily frequency. Competition keeps a floor under how badly you can do.
3. Summer leisure on high-frequency routes. Google's data puts the July/August window at 13 to 44 days out, the shortest of any period it measured.
4. Anything you are holding refundable. This is the point. If you have a flexible booking, you are not waiting. You are holding a floor with an option to improve on it, which is a strictly better position than an empty calendar and a price alert.

What No Guide Tells You About Booking Timing

The flexible rate is priced by professionals. Revenue managers know their cancellation curves better than you know your own plans. Treat the premium as the honest price of uncertainty, not as a loophole.

The saving is often smaller than the hours. We once rebooked a hotel in Seville four times across three weeks. We saved €38 and spent well over an hour doing it, and we were mildly stressed for the whole period. That is a bad wage and a worse holiday. Two checks. Then stop.

Rebooking loses your place in the queue. Your original reservation may carry a room assignment, a note about a quiet floor, a loyalty tier, an early check-in request. Cancel it and you are a new booking at the back of the list, on a night that may now be fuller than when you first booked.

The alert is a marketing channel. Price-drop notifications exist to bring you back to a booking surface. Sometimes the alert is genuinely useful. It is never neutral. Ours would not be either, which is why we would rather tell you to put two dates in a calendar than to install anything.

One of these decisions matters more than the other. For most trips the hotel is the bigger and more movable number, and the flight is the one where hesitating actually costs you. People do it exactly backwards: they refresh the flight daily and book the hotel at the last minute.

Flights vs Hotels: When to Book Each

Where you sleep is the half of this you have the most control over, and the half most worth an hour of thought before you book anything. If you are still choosing a neighbourhood rather than a rate, our guide to where to stay in Athens is the same argument applied to one city.

And where supply is the binding constraint rather than the date — an island with three hotels, a ferry that runs twice a day — the timing advice above stops being the interesting question at all. Our decision guide to the Greek islands is the worked example: five constraints that settle the destination before a price is checked.

And when the date is fixed and the destination is not, the choice moves one level up. Our scored comparison of the best cities in Italy rates twelve of them on cost and crowds alongside the sights, which is the version of this question that a fixed week in August actually asks.

Frequently Asked Questions

Sources and Further Reading

• Expedia, 2026 Air Hacks Report — booking and departure day data, booking windows, seasonal price differences. Based on Expedia.com data from December 2024 to November 2025, plus a January 2026 OnePoll survey of 1,000 US international travellers.
• Google Flights price analysis, summarised in Thrifty Traveler's data analysis and TechCrunch's coverage of the "cheapest time to book" feature — four years of price history, booking windows by destination, day-of-week effects, nonstop vs connecting.
• NerdWallet hotel rate analysis, reported by CNBC — 2,500+ hotel rates collected 2019 to 2021.
• D-EDGE hotel distribution study, reported by PhocusWire — cancellation rates and booking lead times across 680 European properties, 2014 to 2018.
• US Department of Transportation, Aviation Consumer Protection: Refunds — the 24-hour rule, automatic refunds, and the definition of a significant change.
• EU Directive 2011/83/EU, Article 16 — exceptions to the right of withdrawal, including point (l) on dated accommodation and leisure services.
• Gizmodo on why the incognito myth persists, which reports the Consumer Reports test of 372 flight searches, and Thrifty Traveler on searching incognito.
• AltexSoft on fare buckets — how airline inventory classes open and close.