The short answer: yes, but it’s almost never the 20% discount people assume it is — and the way Marriott calculates it means the benefit is worth the least in exactly the situations where travelers expect it to be worth the most.

Here’s what’s actually going on.

The mechanic

Marriott calls it “Stay for 5, Pay for 4.” Redeem points for five consecutive nights at the same property on a single reservation, and one night comes off the bill. No elite status required — every Bonvoy member gets it, and it applies automatically when you’re signed in and searching on Marriott’s site or app.

The rules that trip people up:

  • Points only. Standard redemptions and PointSavers rates qualify. Cash + Points bookings do not.
  • Standard rooms. Upgrade Awards and Nightly Upgrade Awards don’t trigger it. If you redeem for a suite, you get the discount calculated on the standard-room rate.
  • One reservation, consecutive nights. Split the stay across two confirmation numbers and you get nothing.
  • Free night certificates and cash nights don’t count toward the five. You can combine them in the same reservation, but you need five points nights to trigger the benefit.
  • Ten nights gets you two free. The benefit stacks in five-night blocks.

The math nobody explains properly

This is the part that matters. Marriott doesn’t discount by 20%, and it doesn’t zero out an average night. It zeroes out the cheapest night of the five.

Under Marriott’s dynamic pricing, nightly award rates at the same hotel routinely vary by 40% or more across a single week. So the discount you actually receive depends entirely on how flat your five nights are.

Two examples at the same hypothetical hotel.

Flat pricing week: 50,000 + 50,000 + 50,000 + 50,000 + 50,000 = 250,000 Cheapest night removed: −50,000 You pay 200,000. That’s a clean 20% discount.

Volatile pricing week: 50,000 + 50,000 + 50,000 + 35,000 + 60,000 = 245,000 Cheapest night removed: −35,000 You pay 210,000. That’s a 14.3% discount.

Same benefit, same hotel, 6 percentage points of difference. And notice the perverse incentive: the more expensive your Saturday night, the less the benefit helps you proportionally, because the system will always hand you the cheap Tuesday instead.

Travelers routinely plan a Wednesday-to-Sunday stay assuming the pricey weekend night will be comped. It won’t be. The quiet midweek night will.

The one place this gets genuinely interesting

Flip the question. Instead of asking “how much does the fifth night save me,” ask “what does it cost me to add a fifth night?”

Say you’re booked for four nights at 50,000 + 50,000 + 50,000 + 35,000 = 185,000 points.

Add a fifth night priced at 60,000. Your new total is 245,000, minus the cheapest night (35,000), for 210,000. The extra night cost you 25,000 points instead of 60,000 — a 58% discount on the marginal night.

Now add a fifth night priced at 30,000 instead. New total: 215,000, minus the new cheapest night (30,000), for 185,000. Identical to the four-night price. That night is genuinely, entirely free.

This is the actual planning insight. Extending a four-night trip to five never costs the full price of the fifth night, and if the night you’re adding is cheaper than every night already booked, it costs nothing at all. Shoulder nights on either end of a stay are often the cheapest ones on the calendar. Run the search both ways before you book.

The reverse also holds: if your five nights include one unusually cheap night, check whether booking four nights on points and paying cash for the cheap one comes out ahead. Sometimes it does, especially if the cheap night has a low cash rate too.

Does it rescue a bad redemption?

Not really. Bonvoy points are worth roughly 0.8 cents each by NerdWallet’s current valuation — the weakest of the major hotel currencies by a wide margin. A 14–20% discount moves a 0.7-cent redemption to somewhere around 0.8–0.85 cents. It improves a mediocre redemption into an acceptable one. It does not turn a bad one into a good one.

The honest framing: this benefit is a reason to consolidate a stay you were already taking at a hotel you were already going to book with points. It is not a reason to redeem points at a property where the cash rate is cheap.

Two costs that eat into it

Resort and destination fees. Marriott does not waive them on award stays, and elite status doesn’t help. A five-night award at a property charging $50 a night in destination fees costs you $250 in cash on a “free” stay. Fold that into your math before you decide the fifth night made the redemption worthwhile — it frequently cancels out the entire discount and then some.

No refunds on the unused portion. If you book five nights and check out after four, Marriott treats the night you skipped as the free one. You don’t get points back for leaving early. If your plans are genuinely uncertain, the safer play is booking fewer nights and extending later, accepting that you’ll lose the discount if the extension doesn’t land you back at five.

The verdict

Worth using: nearly always, when you were already booking five-plus consecutive points nights at one property. It’s free money and it applies automatically.

Worth planning around: sometimes. Specifically, when nightly award pricing is flat, or when you’re deciding whether to add a shoulder night to a four-night stay.

Worth stretching a trip for: rarely. If you’re adding a night purely to trigger the benefit, you’re spending points and a day of your life to save 14–20% on a currency worth 0.8 cents. Add the night because you want the night.

Before you book, do this: pull up the five-night award total, then pull up the four-night total, and look at the difference. That number — not the advertised 20% — is what the benefit is actually worth on your specific dates.