How to Find Cheap Flights.

You find a price you like on Monday. Thursday you come back to book it and it's gone up by $180. Incognito mode, because you read somewhere that this helps. You're not entirely sure it does anything. You check again. The price is now different from both previous numbers and you have no idea why.

View larger

This isn't bad luck or a browser quirk. It's just how flight pricing works, and most cheap flight guides don't acknowledge it. They give you a list of rules. The same list, rephrased, that every other article about cheap flights has been recycling for a decade. Book six weeks out. Fly on Tuesdays. Clear your cookies. Some of those rules have some truth in them. Some are outdated. Most are simplifications of a system that doesn't actually operate on rules.

What actually helps is understanding what's behind the price. Not so you can game it. The pricing algorithms airlines run today are too complex and too opaque for consistent gaming. The point is to stop searching in ways that don't improve your position and start focusing on the inputs that shift things.

How Flight Prices Actually Work

Airlines price every seat by algorithm. The cost of a seat changes constantly based on demand, how many seats remain in each pricing tier, competitor pricing on the same route, time to departure, and factors that aren't publicly disclosed. There is no fixed moment when fares are reliably cheap. There is no universal booking window that unlocks lower prices across every route.

What there are instead are patterns. They hold up, but only probabilistically. The same seat on the same flight might cost $280 when you search on a Tuesday morning and $440 on Friday afternoon, for reasons that have nothing to do with your search history or whether you've looked before. Prices fluctuate because demand signals fluctuate: a corporate event books out business class on your route, a sale on competing routes spills demand onto yours, a holiday weekend fills Friday flights while Monday stays empty. These are systemic movements, not responses to you personally.

One thing follows directly from that and almost nobody does it. Seats are sold in buckets, and a search for four passengers has to find four seats in the same bucket or it prices all four at the next one up. So when a fare looks oddly high for a group, search it as one passenger.

If the single-seat price is much lower, the cheap bucket has two or three seats in it and you are being quoted the expensive one for all of you. Book those seats on separate reservations and pay the low price for the ones that exist. The catch is that separate reservations are not linked, so if the flight moves, you are four separate problems rather than one.

Understanding this is useful because it changes your posture. You're not hunting for the right trick. You're managing four variables: flexibility, timing, routing, search tools. All of them consistently shift your odds of landing a better price. None of them guarantee anything. All of them matter.

Flexibility Is the Most Valuable Thing You Have

Before any search technique, any app, any alert system: the single biggest factor in how cheap your flights can get is how flexible you are. Not slightly flexible. Structurally flexible in one or more of the ways that actually affect pricing.

Flexibility on dates

Flexibility at the date level means you're willing to shift departure or return by a few days, or search across a full week rather than a fixed date. Tuesday and Wednesday departures generally price under Friday and Sunday ones, which fill with weekend travelers, business commuters, and people catching up on delayed plans. Saturday morning often undercuts Saturday midday. These are tendencies rather than guarantees, and published estimates of how large the effect is disagree by an order of magnitude, which is itself the answer. Search a full week and the range on your own route shows up.

Flexibility on routing and airports

Flexibility at the routing level means you're willing to consider a connecting flight, an alternate departure airport, or a different city as your entry point to a destination. Flying into a secondary airport and taking a short train or bus adds a few hours but can take a serious bite out of the fare. It depends entirely on the transfer. A secondary European airport plus a €20 train into the city you actually want is usually worth it. The same distance by $100 taxi is not.

If you're within driving or train distance of more than one departure airport, checking both consistently surfaces price differences. Markets with more route competition have lower prices. A one-hour train to an airport with better carrier competition can save $150 or more on a transatlantic fare, which usually makes that extra hour worth it. Price the whole door to door leg before you call that a saving. Parking at the far airport, the train fare, and an early night in a hotel you would not otherwise book are the hidden travel costs that erode the saving.

Flexibility at the destination level is the deepest kind, and the one with the most upside. If you're deciding where to go rather than locked into a specific place, searching by region and price opens up a different category of options, and most major flight search tools have a feature for it. Use it early, while you're still orienting, before you've committed to a country let alone a city.

The Timing Question

The advice to book six weeks out comes from research showing that fares tend to dip before rising sharply as departure approaches. That pattern holds for some routes. For popular summer transatlantic routes, or Southeast Asia at Christmas, prices typically rise from the moment seats open and don't return to the early-booking level. For less competitive routes between secondary cities, fares can be stable for months and the six-week window is largely meaningless.

View larger

What holds up more consistently: booking early on routes where demand is predictably high. Airlines release seats in pricing tiers. The lowest-priced ones stay available until that tier fills and closes, and then the next tier opens higher, and the one after that higher again. Get in early enough on a route you know will fill and you catch the cheap tier before it disappears. Wait until two weeks out and you pay whatever is left. On a popular route in peak season, several hundred dollars more.

The window two to four weeks before departure is generally the most expensive period on most leisure routes. Business travelers book here. People who've decided late book here. Airlines price accordingly. If you're in this window and the price feels high, it probably is. Booking earlier next time is the main answer.

The price feels high because it is high. You are in the wrong booking window for the kind of trip you are planning.

Last-minute deals exist, mostly on budget carriers that would rather fill a seat at a reduced price than fly it empty. This strategy works occasionally for short-haul routes on slower travel days. It falls apart during busy periods, doesn't function well on routes with limited competition, and creates problems when accommodation, tours, and other bookings need to align with a specific date.

For most independent travelers: book popular or peak-season routes three to four months out. For more flexible or less competitive routes, a one to two month window often balances certainty against price. Avoid the final three weeks unless you have no other option.

The calendar does some of this work for you before any of the above applies. January and February are the lowest-fare months of the year, with September and October close behind, and July and December are the two that punish you. That is a bigger lever than any search technique on this page, and it is free if your dates are not fixed by somebody else.

Time of day is the one that will not die. Red-eyes and six in the morning departures are widely believed to cost less, and the airlines that publish on this call it a leftover from when fares were loaded by hand each morning.

Sometimes the early flight costs less because fewer people want it, which is demand rather than the hour. Buy the awkward departure because it fits your day or because the fare in front of you is lower, not on principle, and price the other side of it first: what jet lag does to the first two days is what a night flight actually costs you at the far end.

Routing: The Part Most People Skip

Where you fly from and through often matters more than the specific booking window. It gets less attention than it deserves. Direct flights carry a price premium on most routes, not always, but consistently enough to make the comparison worth running. A one-stop routing that adds three hours often comes in well under the nonstop, and the gap is usually wide enough to change the decision rather than to round it.

If you're already spending a full day in transit either way, the tradeoff usually favors taking the stop. Three hours is not a day. What a stop does change is the state you arrive in, which the fare never shows and the jet lag side of the same decision does.

The hub matters. Some connecting airports cost substantially less to route through than others, based on how much carrier competition exists at that hub. Routing a long-haul flight through a hub with three competing airlines rather than one dominant carrier can produce meaningful price differences. The search engines surface these combinations when you leave options open, but you occasionally need to try a few manual variations to find what's actually available at the lower end.

On the arrival side: many destination cities have more than one airport, or a neighboring city served by more routes, which usually means a lower fare, with good onward connections. The search engines don't always surface these alternatives when you've typed in a specific destination.

If you're not finding prices you like, trying the secondary airport manually takes five minutes and occasionally moves the price a long way. It can also move against you without the fare showing it. Getting to El Nido is a worked case: a whole domestic network moved out of one Manila airport and into another two hours north, and the ticket price never mentioned the two hours.

Two more routing moves worth trying before you settle. The first is pricing two one-way tickets against the round trip. On short-haul and inside low-cost networks the pair often wins, and it lets you fly out on one carrier and home on another, which is how you get the cheap outbound without accepting the terrible return attached to it. It does not always win, and on long-haul it frequently loses badly, so price it rather than believing it.

The second is asking whether you are a category. Several carriers run youth or student fares for roughly the eighteen to twenty five band and some run senior fares at the other end, and the discounts run from token to substantial depending on the route and how late you book. They are almost never surfaced in a normal search. You have to look for them on the airline site or ask.

The cheapest route is usually not the route you asked the search box about.

Two more that live one layer down. The open jaw is the round trip that does not come back to where it started: fly into one city, out of another, and let the ground in between be a train ride you were going to take anyway. Booking engines sell it under multi-city and it frequently prices under two separate round trips.

The positioning flight is the uglier cousin. When the fare from your own airport is bad and the fare from a bigger one two hundred miles away is not, you buy the cheap seat to the bigger airport separately and begin the trip properly from there. It works because carriers price against the competition at that airport rather than against where you happen to live.

Two tickets that do not know about each other will not wait for each other either, so leave a night in between rather than an hour. And if the airline whose hub sits between your two ends sells a stop there as a product rather than as a penalty, look at it before you write the connection off as dead time. On a few transatlantic carriers the break costs nothing on the airfare and turns a layover into a second city.

Search Tools Worth Using

Google Flights, Skyscanner, and Kayak all pull from overlapping data sources but present them differently and occasionally surface different combinations. Using more than one for a trip that matters takes extra time and is worth it.

Google Flights has the most useful interface for flexible searching. The calendar view shows price variation day by day across a full month, so a whole month of prices is one glance instead of thirty searches. The Explore feature searches by region and price from your home airport, which is useful when you haven't committed to a destination. It is fast and does not bury you in upsell prompts.

Skyscanner is particularly useful for long-haul routes and for finding combinations of carriers that a single airline's site wouldn't show. The cheapest month view is blunt but effective for identifying which part of a month to focus on. It also handles budget carriers better than Google Flights on some regional routes.

Kayak has a solid price alert system, which matters when you're tracking a specific route over time rather than searching once and deciding. For routes where you're willing to monitor prices across several weeks, setting an alert and checking it periodically is less stressful than daily manual searching.

One principle worth keeping: use search engines for discovery, then book directly with the airline when you've found a price worth taking. The stronger reason is what happens when you need to undo it: the large agencies generally run their own 24-hour cancellation and the smaller ones surfaced through comparison engines generally do not, so the rule turns on which agency rather than on agency against airline. Some airlines also charge booking fees on third-party sites, and direct booking gives you better access to manage the reservation if plans change, which matters more when you've paid for a flexible fare.

On budget carriers it avoids added fees that erode the price advantage you found in the first place. Discovery in one place, purchase in another. Which is also the fix for the blind spot the three engines share: some carriers do not sell through them at all, and the ones that do are not always fully loaded, so a route with no result on Google Flights is not a route with no flights. If a destination looks unreachable, find out who actually flies there and open that airline directly.

Booking direct also buys you a federal rule, and it is not quite the rule people repeat. On flights to, from or within the United States, a carrier has to let you either hold the quoted fare without paying for twenty four hours, or cancel without penalty inside twenty four hours, as long as you booked at least a week before departure. The word between those two is or.

The carrier picks which one it offers and does not have to offer both, so find out which one you are being given before you rely on it. Where the airline chose cancellation, that day is the cheapest research window in the whole process: book the fare that is about to move, then spend the evening checking whether it was actually the good one.

One thing to watch on the airline site itself: which currency it decides you are buying in. A ticket priced in the local currency of the route can come out several percent under the same ticket priced in yours, and the terminal or the checkout will then offer to convert it back for you, which is the expensive direction. Pay in the currency the airline quotes, and use a card that does not charge for that. What your money actually costs abroad is the same screen and the same trap, on the ground.

Flight Alerts: When They're Worth Setting

Price alerts are useful in a specific situation: you have a route in mind, you have genuine flexibility on timing, and you'd like to be notified of price changes rather than checking manually every few days. They're most valuable when they are already running before your booking window opens rather than inside it: six months out for a peak long-haul summer, a couple of months for a domestic hop.

They work best when you have a benchmark. If you've done enough searching to know that $420 is a good price for your route and the current price is $510, setting an alert at $450 means you'll be notified when prices move meaningfully in the right direction. An alert set without any sense of what a normal price looks like is just a notification without context, which doesn't help you decide whether to act. If you don't have that number yet, build a benchmark with a simple travel budget before you set anything.

Building that benchmark takes about ten minutes and three searches. Open the route with flexible dates and read the price calendar across a whole year, not a week. Note the floor, which is what the route costs when nothing is happening, and note the ceiling in the school holidays. Then search it once more for the dates you actually want. Now you have three numbers: the floor, the ceiling, and yours, and the only useful question is which of the three your quote sits nearest. A number somebody else published for a different origin is not a benchmark. Yours is.

A price is only high or low against a number you already have. Without one you are not deciding, you are reacting.

The main failure mode with price alerts is using them as a substitute for booking. Tracking becomes its own project. Three months pass. The fare you were watching disappears. A useful question to break out of this: if the price went up $60 tomorrow, would I regret not having booked today? If yes, and the price is reasonable for the route, book it. Prices can go both ways, and the mental overhead of monitoring a specific flight across months adds up and often is not worth it.

What Doesn't Work Much

View larger

Incognito mode and clearing cookies. The mechanism by which your browsing history would raise your individual flight price as you browse isn't clearly established. Airlines and booking sites deny doing it, and multiple independent tests have failed to find consistent evidence for the effect. The price variations people attribute to their own search activity are almost certainly the normal churn of a pricing engine that moves for reasons that have nothing to do with you.

Always flying budget carriers. Low-cost airlines often cost less, not always. On some routes, full-service carriers with comparable base fares offer meaningfully better baggage allowances, more flexibility on changes, and less aggressive fee structures for when things go wrong. Running the full comparison including checked bag fees, seat selection, and change costs gives you the landed price, which sometimes flips the apparent result. Bag fees are the line that does the flipping, because of how they count.

A checked bag on a major US carrier runs about $45 prepaid and about $50 at the airport, and it is charged per person, per direction. Budget carriers charge more than that and charge for the carry-on as well, sometimes triple at the gate what it cost at booking. Two people with a bag each on a return trip is four bag charges before anybody has chosen a seat, which is most of the gap you thought you were saving.

Waiting indefinitely for a flash sale. Major sale events still happen, but they're less predictable and less dramatic than they were a decade ago. Planning a big trip around the hope of a future price drop is poor strategy for anything with a fixed travel window. If you find a price that seems reasonable and you'd accept a $50 increase without hesitation, book it rather than gambling on something better appearing.

Error fares are the exception people point at, and they are worth understanding rather than waiting for. A mistake in a currency conversion or a fuel surcharge occasionally puts a long-haul ticket on sale for a fraction of its price, and those fares get found within hours. You cannot plan a trip around one. You can be ready to take one: know where you would go, keep the dates loose, and book fast, because airlines sometimes cancel the booking and sometimes honor it, and neither is owed to you.

And then there is the trick that keeps coming up, which is buying a ticket to a farther city and getting off at the connection because the longer route prices lower. It works, right up until it does not. You cannot check a bag, because the bag goes where the ticket goes. One direction only, since the rest of the itinerary cancels the moment you skip a leg. And doing it repeatedly is a breach of almost every airline contract, which has cost people their miles and their accounts. Not illegal. Just not free, and the price is charged later.

• Daily monitoring of a single route for months on end
• Clearing cookies obsessively before each search
• Refreshing one tab while expecting the price to drop
• Booking a lower fare that misconnects through an airport you don't know

That last one deserves a sentence of its own, because it is the way a cheap fare turns expensive. If the two legs are on one ticket, a missed connection is the airline problem to solve and they rebook you. If they are on separate tickets, which is what a self-transfer is and what the search engines will happily sell you, nobody owes you anything.

The second airline sees a no-show. You buy a new ticket at the walk-up fare, which is the most expensive ticket in the building. Leave three hours on a domestic self-transfer and four to six on an international one, and if the plan needs a bag rechecked and an immigration queue crossed, treat that as the floor rather than the target.

When You Arrive

One cost that flight guides consistently skip: data on arrival. Getting from the airport to your accommodation in an unfamiliar city almost always requires functioning maps and the ability to send a message. Airport SIM kiosks are slow and expensive. Standing in a queue after a long-haul flight, trying to get a SIM card working while carrying everything you own, is a specific kind of friction you don't need at the beginning of a trip.

An international eSIM set up before departure solves this cleanly. You activate it as you land and have data before you're out of the terminal. A travel eSIM covers most international destinations and is straightforward to configure before you leave home. Small cost. It removes one friction point from the worst part of any trip, the first disoriented hour in a new place. If you're weighing that against a local SIM, the full eSIM vs physical SIM comparison sets the two side by side.

Cheap Flights and Travel Insurance

There's a structural tension in cheap flight searching that most advice skips over. The cheapest tickets are usually the least flexible. Non-refundable budget fares with no rebooking rights. Advance purchase tickets that evaporate if your plans change. Multi-leg itineraries routed through airports where a delay on the first leg becomes entirely your problem on the second. Season does the same thing from the other side. The cheapest months to fly somewhere are usually the months something is wrong with the weather there, and the Philippines in the rainy season is what that trade actually costs you once you have landed.

One right survives all of it, and the fare class cannot sell it away. If the carrier cancels the flight, or delays or changes it significantly, and you decide not to fly it and not to take the voucher, a nonrefundable ticket on a route to, from or within the United States has to be refunded in full, taxes and ancillary fees included, in the form you paid with.

The cheapest seat on the aircraft gives up the right to change your mind. It does not give up the right to be repaid when the airline changes theirs. Which is worth knowing before you accept the credit they will offer you first.

The more you strip out of the base ticket price, the more valuable some protection becomes for what you can't predict: illness before departure, a family emergency, weather delays that cascade through your connections. Travel insurance doesn't have to be expensive, but it's worth reading what a policy actually covers rather than assuming the cheapest option does the same job as one that covers more.

A policy that excludes the scenarios you're most likely to actually face is not a saving. It's a false economy. Sorting this out before you book, rather than after, is generally the right order of operations. Start with what travel insurance actually covers, then compare policies against your own trip. Rolling travel medical insurance is one of the options built for open-ended trips, where the policy renews monthly instead of against a fixed pair of dates.

What the Process Actually Looks Like

In practice, searching well for cheap flights looks less like a system and more like a few deliberate sessions. You have a trip in mind, or a rough idea of one. An hour in Google Flights, checking dates and destinations against prices with as much flexibility as your situation allows. Note what looks reasonable. Check a second search engine for comparison. Set an alert if you're not ready to commit.

You come back once or twice over the next few weeks. The price moves slightly or holds steady. At some point you have enough of a benchmark to recognize a fair price when you see it, and you book.

No daily monitoring. No browser tricks. Nothing obsessive about tracking a price that may never reach the number you've decided is the right number. Just a clear understanding of what you're looking at, genuine flexibility where you can build it in, and enough patience to wait for the right timing on routes where timing matters.

Prices will still surprise you. Both directions. A fare you were watching drops $180 on a Wednesday you weren't checking. Something you booked three months out turns out to cost the same the week before departure. That cheap connecting itinerary misconnects and costs you a full day of the trip.

This is the system. It isn't gameable. But it is navigable, and there's a wide gap between people who understand the mechanism and people working off the same recycled tip sheet. The goal was never the cheapest possible flight. It was always a fair price, found without spending more of your life on it than it deserves.