How far in advance should I book a hotel?

It depends on whether supply is constrained on your dates. For an event, a peak season with limited supply, or a specific room type such as a family room, book as soon as your dates are firm. For an ordinary stay in a city with plenty of rooms, the last one to two weeks is often when prices soften.

Do hotel prices go down closer to the date?

They can, unlike flights. An unsold room earns nothing, so hotels with unsold inventory discount close to arrival. This only happens where supply exceeds demand - during events and peak seasons prices rise instead.

Is there a cheapest day of the week to book?

Not reliably. Which nights you stay matters far more than which day you book. Business cities are expensive midweek and cheap at weekends; leisure destinations are the reverse.

What is the safest strategy if I am not sure?

Book a refundable room as soon as your dates are firm, then check again about two weeks before arrival. If the price has fallen, book the new rate first, confirm it, and only then cancel the original before its deadline.

Are last-minute hotel deals real?

Yes, in markets with unsold rooms, and they can be substantial. They are not real during events or peak season, when late booking simply means paying more for whatever is left.

When to book a hotel: how far ahead, and when prices actually drop

The advice to "book early" is half right. What matters is which of two pricing regimes your trip falls into.

There is no universal best day or best number of weeks ahead. Anyone who gives you one is selling something. What there is, is a small number of mechanisms that reliably drive hotel prices, and once you know which one applies to your trip, the timing follows.

Hotels do not price like airlines

This is the misconception underneath most bad advice. Airline seats are perishable in a fixed cabin with well-established yield curves, and fares generally climb as departure nears. Hotels have a different shape, because a hotel can — and does — cut prices close to arrival to fill rooms that would otherwise earn nothing.

An empty room tonight earns zero forever. A hotel with 30 unsold rooms at 4pm has a strong incentive to discount. This is why last-minute hotel deals are real in a way last-minute flight deals largely are not.

The two regimes

Almost every trip falls into one of these, and they point in opposite directions.

Constrained demand: book early, and early means early

When the number of rooms in a place is genuinely limited relative to demand on your dates, prices only go one way. This is the regime when:

there is a major event — a conference, a festival, a match, a religious pilgrimage;
it is a peak season in a destination with limited supply (a small island in August, a ski town at New Year);
you need something specific — a family room, adjoining rooms, accessible facilities, four beds. Specific inventory is thin and sells first regardless of season.

Here, waiting does not get you a discount. It gets you a worse hotel further out, at a higher price. Book as soon as your dates are firm, on a flexible rate if one exists.

Unconstrained demand: the last two weeks are yours

When supply comfortably exceeds demand — a large business-hotel city on a weekend, a resort in shoulder season, any big city outside an event — the pressure runs the other way. Hotels approaching arrival with unsold rooms discount them.

Here, booking six months early buys you nothing. The last one to two weeks is where the softness is.

How to tell which one you are in

Search your dates now, and search the same weekday a month later. If your dates are dramatically more expensive than the comparison week, you are in a constrained period and should book. If they are similar, you have time.

Where the real savings are, and they are not in timing

Two decisions move the price more than the booking date does.

Which nights. In most business cities, Tuesday and Wednesday are the expensive nights and the weekend is cheap. In leisure destinations it is exactly inverted. Shifting a two-night stay by one day frequently saves more than a month of watching prices.

Which week. The week either side of a school holiday, a public holiday or a major event is often a fraction of the price of the event week itself, in the same hotel, with the same weather.

The flexible-rate strategy

If the hotel offers free cancellation, the optimal play is straightforward:

Book a refundable room as soon as your dates are firm. You now have a guaranteed room at a known price and no downside.
Check again about two weeks out.
If it has dropped, rebook and cancel the original before its deadline.

You capture the discount if it appears and you are protected if it does not. The only cost is remembering to check — and the only real risk is cancelling the first booking before the second is confirmed. Confirm the new one first, always.

Do prices drop at the very last minute?

Sometimes, and it is a genuine gamble rather than a strategy. Same-day rates in a soft market can be excellent. In a constrained market, same-day means whatever is left, which is the worst room in the worst hotel at a price that does not reflect either.

Do this only if you can genuinely sleep somewhere else if it fails.

What we can measure, and what we cannot

Broad claims like "book on a Tuesday, 43 days out" are the kind of statistic that circulates widely and survives no scrutiny — the effect is small, market-specific, and does not persist.

Where we hold enough real bookings for a city to say something meaningful, we publish it in our Travel Index with the sample size attached, so you can see exactly how much evidence sits behind it. Where we do not hold enough, we publish nothing rather than a number that sounds authoritative and is not.

When to book a hotel: how the booking window actually affects the price