These are simple enough that the arithmetic is never the problem. What causes trouble is what goes into them, which is why the definitions below are more specific than the formulas.
Occupancy = rooms sold ÷ rooms available ADR (average daily rate) = room revenue ÷ rooms sold RevPAR (revenue per available room) = room revenue ÷ rooms available, or equivalently occupancy × ADR ADR divides by rooms *sold*; RevPAR divides by rooms *available*. That single difference is the entire reason both exist, and mixing them up produces a number that looks plausible and means nothing.
A worked example A 200-room hotel sells 150 rooms on a given night for $22,500 in room revenue.
Occupancy is 150 ÷ 200 = 75%. ADR is $22,500 ÷ 150 = $150. RevPAR is $22,500 ÷ 200 = $112.50, which is also 0.75 × $150.
Now compare with a night where the same hotel sells 180 rooms for $23,400. Occupancy rises to 90% and ADR falls to $130 — a worse rate, and a manager looking only at ADR would call it a bad night. RevPAR is $117, which is higher. The second night made more money from the same building, which is exactly what RevPAR exists to show.
The denominator mistake The most common error in this arithmetic is quietly removing rooms from "rooms available". A wing is closed for refurbishment, or eight rooms are out of order, and the occupancy calculation starts dividing by 192 instead of 200 because that is what was really sellable.
This is defensible for judging the commercial team and indefensible for judging the asset. Excluding out-of-order rooms makes occupancy and RevPAR rise the moment rooms break, which means the metric improves as the building deteriorates. Whichever convention you adopt, use one, state it on the report, and never switch mid-year — most of the alarming or miraculous trends in hotel reporting are a denominator that changed without anyone saying so.
Convention Denominator Occupancy RevPAR at $22,500 Physical rooms 200 75.0% $112.50 Sellable rooms 192 78.1% $117.19
Same night, 200 physical rooms, 8 out of order, 150 sold What RevPAR does not tell you RevPAR is a revenue measure with no notion of what the revenue cost to earn, so it can be pushed up by discounting into a full house that loses money once housekeeping, laundry, amenities and utilities are counted. The last twenty rooms sold on a heavily discounted night are frequently the least profitable rooms of the month.
Two habits guard against this. Watch RevPAR alongside a profit measure — GOPPAR (gross operating profit per available room) if you have it — rather than on its own. And pair every occupancy forecast with the staffing it implies, because the variable cost of a high-occupancy night is mostly housekeeping hours, and that is the number a rate decision is really trading against.
Use the forecast, not the average The reason to compute these daily rather than monthly is that the operational decisions they drive are daily. An average occupancy of 78% tells you nothing about how many housekeepers to roster for Thursday.
The chain worth building is: forecast occupied rooms, split into checkouts and stayovers, convert to housekeeping workload, and staff to that. Monthly averages are for reporting to owners; daily forecasts are for running the hotel, and confusing the two is how a property ends up correctly staffed on average and wrongly staffed every day.
Common questions What is the formula for RevPAR? RevPAR equals room revenue divided by rooms available, which is mathematically identical to occupancy multiplied by ADR. A 200-room hotel earning $22,500 in room revenue has a RevPAR of $112.50, whether or not every room was sellable that night.
What is the difference between ADR and RevPAR? ADR divides room revenue by rooms sold, so it measures the rate achieved on the rooms that were actually let. RevPAR divides by rooms available, so it measures how well the whole building earned. A night with a lower ADR can have a higher RevPAR, and usually made more money.
Should out-of-order rooms be included in occupancy? Either convention is defensible, but excluding them makes occupancy and RevPAR rise as rooms break, which flatters the asset as it deteriorates. Whichever you choose, state it on the report and never change it mid-year — a shifting denominator explains most surprising trends in hotel reporting.
What does RevPAR not measure? Cost. RevPAR is a revenue figure with no notion of what earning it required, so it can be raised by discounting into a full house that loses money once housekeeping, laundry and utilities are counted. Pair it with a profit measure such as GOPPAR.