Hotel occupancy in the United States has largely recovered from the collapse triggered by the pandemic, but the industry has yet to return to the levels seen immediately before it.

U.S. hotels recorded an average occupancy rate of 62.3% in 2025, down from roughly 63% in 2024, according to year-end data from CoStar. It was the first annual decline in hotel occupancy since 2020. For 2026, CoStar and Tourism Economics expect occupancy to improve to 63.1%, supported by stronger room demand.

That would represent another step toward normalization, but it would still leave the industry below the 65.8% occupancy rate recorded in 2019.

Hotel occupancy is one of the simplest measures of lodging demand, but understanding exactly what it represents is important.

STR, which is part of CoStar, defines occupancy as the percentage of available hotel rooms sold during a specified period. The calculation divides rooms sold by rooms available. Rooms sold exclude complimentary rooms, while room supply reflects the number of available rooms multiplied by the number of days in the period.

A 70% occupancy rate therefore means that hotels sold 70 out of every 100 available room nights during the period being measured. It does not mean that 70% of every hotel room in the country was occupied on every individual night.

That distinction matters because national occupancy combines thousands of properties across different markets, hotel classes, seasons and days of the week.

STR collects information including rooms available, rooms sold and net room revenue from participating hotels on daily, weekly and monthly schedules. CoStar said its global hotel performance sample covered approximately 95,000 properties and 12.2 million rooms as of October 2026.

The industry's recent history shows both the scale of the pandemic disruption and the speed of the subsequent recovery.

Historical figures through 2023 come from Oxford Economics and STR data published by the American Hotel & Lodging Association. CoStar's year-end data put 2024 occupancy at approximately 63.0%, followed by 62.3% in 2025. The 2026 figure is a forecast from CoStar and Tourism Economics rather than a completed full-year result.

The table also illustrates why forecasts should not be treated as reported results. An earlier AHLA report, using Oxford Economics and STR data, had projected occupancy of 63.38% for 2025. When the year was complete, CoStar reported an actual industry occupancy rate of 62.3%.

The most dramatic movement occurred between 2020 and 2022.

Average U.S. hotel occupancy fell to 43.89% in 2020, compared with 65.80% in 2019. By 2021, it had recovered to 57.50%, and by 2022 it had climbed above 62%.

The pace of improvement then slowed considerably. Occupancy increased only modestly from 62.54% in 2022 to 62.97% in 2023 and approximately 63% in 2024.

That pattern marked a shift from recovery toward stabilization. The large amount of hotel demand that returned after the pandemic had already been restored, leaving the industry closer to its normal relationship between room supply and travel demand.

Even so, occupancy remained below its pre-pandemic level. The gap between 2024's approximately 63% rate and 2019's 65.8% amounted to nearly three percentage points.

The recovery lost some ground in 2025.

CoStar reported a full-year occupancy rate of 62.3%, representing a 1.2% decline from 2024 on its year-over-year comparison. Average daily rate, or ADR, still increased 0.9% to $160.54, while revenue per available room, known as RevPAR, declined 0.3% to $100.02.

Occupancy, ADR and RevPAR measure different parts of hotel performance. Occupancy measures the proportion of available rooms sold. ADR measures the average room revenue generated for each room sold. RevPAR divides room revenue by all available rooms, combining the effects of both occupancy and pricing.

A hotel industry can therefore experience weaker occupancy while maintaining or increasing room rates. That is essentially what happened nationally in 2025, when ADR rose even as occupancy declined.

Hotel demand improved during 2026, prompting CoStar and Tourism Economics to raise their outlook.

In August, the organizations increased their full-year occupancy forecast to 63.1%, 0.3 percentage points above their previous projection. CoStar said U.S. hotels sold a record number of room nights during the first half of 2026, with room nights sold increasing by 11.4 million from the same period of 2025. Room revenue increased by more than $5.4 billion.

Monthly results also showed stronger demand. U.S. hotel occupancy reached 69.6% in June 2026, up 1.6% from June 2025, while August occupancy stood at 66.4%, 0.5% above the previous year.

These monthly figures should not be compared directly with annual occupancy rates. Hotel demand is highly seasonal, with summer vacations, business travel, conventions, holidays and major events causing occupancy to rise and fall throughout the year.

A national average also conceals substantial differences between individual hotel markets.

In 2025, New York City recorded an 84.1% occupancy rate, the highest among the markets included in CoStar's Top 25 Market comparison. Its average daily rate reached $333.71, while RevPAR was $280.71.

Other markets moved in different directions. St. Louis recorded the strongest year-over-year increase in occupancy among the Top 25, rising 6.5% to 61.7%, while Houston posted the largest decline, falling 8.6% to 58.9%.

These figures describe performance within CoStar's Top 25 Market group rather than a comprehensive ranking of every U.S. city or hotel market.

Shorter periods can produce even wider differences. In August 2026, San Francisco reported occupancy of 79.3%, while New Orleans recorded 43.3%. Those are monthly readings, however, and should not be interpreted as annual occupancy levels.

Weekly hotel data provide a more immediate view of demand, but they require even more caution.

For the week of October 11 through October 17, 2026, U.S. hotel occupancy reached 70.7%, 3.1% above the comparable week in 2025. ADR was $177.65 and RevPAR reached $125.53.

A weekly rate above 70% does not mean annual occupancy is moving toward that level. Conferences, holidays, sporting events and differences in the calendar can materially change a single week's results.

CoStar's weekly reports regularly identify such effects. For example, some 2026 comparisons were influenced by the timing of Labor Day and Rosh Hashanah, while major events lifted individual markets during other weeks.

Annual averages smooth out much of this volatility and are therefore more useful when examining the industry's broader direction.

A higher occupancy rate is generally associated with stronger room demand, but it does not by itself show whether hotels are becoming more profitable.

Hotels can improve revenue by charging higher rates even without filling more rooms. Conversely, a property can achieve high occupancy by discounting heavily, which may not produce the strongest financial result.

This is why the hotel industry commonly evaluates occupancy alongside ADR and RevPAR rather than relying on occupancy alone.

The recent U.S. experience illustrates that relationship. Hotel occupancy in 2025 remained well below 2019, yet average room rates were considerably higher in nominal dollar terms. CoStar reported ADR of $160.54 in 2025, compared with $131.56 in 2019 in the historical data published by AHLA. Those figures are nominal and are not adjusted for inflation, so they do not represent the same increase in purchasing power or profitability.

The broad picture is one of substantial recovery followed by slower progress.

Occupancy climbed from less than 44% in 2020 to approximately 63% by 2024, before declining to 62.3% in 2025. Stronger demand in 2026 has improved the outlook, and CoStar and Tourism Economics forecast a 63.1% full-year occupancy rate.

Even if that forecast is reached, U.S. hotel occupancy would remain roughly 2.7 percentage points below the 65.8% recorded in 2019.

The industry's room demand has largely recovered from its historic disruption, but the national occupancy rate shows that recovery and a full return to pre-pandemic utilization are not quite the same thing.