Tracking the estimated settled ADR of Tokyo’s business hotels and city hotels from June to September 2026 (year-on-year, finalized month against finalized month) shows that September marked a turn. Business hotels’ finalized September came in at ¥13,364 (N=914 properties), 2.2% above September 2025’s ¥13,072 (N=894 properties), ending three straight months of year-on-year declines that began in June (June −4.7%, July −5.3%, August −8.9%). City hotels, at ¥24,650 (N=110 properties) against ¥27,467 (N=107 properties), are still in negative territory at −10.3%, but the decline has halved from August’s −20.6%. Estimated OCC (OTA listed-inventory basis) for the same September was 90.3% for business hotels and 90.0% for city hotels, both in the 90% range. This article examines what happened at this seam between summer and autumn, when occupancy held and only the year-on-year rate gap moved, and what it implies for rate design in October and November.

Scope: Tokyo business hotels (estimated settled ADR N=914 properties; estimated OCC N=538–561 properties / 78,348–80,960 rooms) and city hotels (estimated settled ADR N=110 properties; estimated OCC N=52–54 properties / 16,054–17,294 rooms). The price metric in this article is estimated settled ADR (the settled price level estimated from OTA and other sales data, tax-excl. equivalent); occupancy is an estimate based on OTA listed inventory. Both are defined at the end of the article. Data as of: October 25, 2026.

Finalized June–September YoY — business hotels turned in September, city hotels halved their decline

First, line up the four months that can be compared on a finalized-to-finalized basis. Because past months of estimated settled ADR are all on a finalized basis, the same month in 2025 and 2026 can be compared directly. Business hotels widened their year-on-year gap into the summer — June ¥11,890 (prior year ¥12,474, −4.7%), July ¥12,306 (¥12,990, −5.3%), August ¥11,554 (¥12,689, −8.9%) — then in September came in at ¥13,364, ¥292 above last year’s ¥13,072. Given that January–May were all positive year on year (+5.1%, +0.4%, +2.9%, +3.9%, +5.7%), the June–August declines read less as a full-year trend than as a phase confined to the summer.

City hotels, by contrast, stand out for both the depth of the decline and the lateness of the turn. The year-on-year gap widened over three months — June ¥22,195 (prior year ¥25,441, −12.8%), July ¥22,434 (¥27,055, −17.1%), August ¥21,265 (¥26,769, −20.6%) — and September, at ¥24,650, is still ¥2,817 short of last year’s ¥27,467. The decline did, however, nearly halve from −20.6% in August to −10.3% in September, so in direction city hotels are riding the same “recovery into early autumn” as business hotels. City hotels’ January–May ran +8.3%, +4.0%, +5.8%, +0.5%, −3.8%: growth stalled in April and turned negative in May, so city hotels entered the year-on-year decline one month earlier than business hotels.

Estimated settled ADR (tax-excl. equivalent; finalized month vs. finalized month). N= is the number of properties estimated in each month. Source: MetroEngines Research; compiled by HotelBank Editorial Team

Averaged over the four months, business hotels were ¥12,279 against ¥12,806 (−4.1%) and city hotels ¥22,636 against ¥26,683 (−15.2%). That the summer decline was concentrated in city hotels is also confirmed by the city-to-business rate multiple. In September 2025 city hotels were 2.10x business hotels; in September 2026 the multiple had narrowed to 1.84x, and August had compressed by the same margin, from 2.11x to 1.84x. In other words, this summer’s Tokyo was a market where business hotels broadly held last year’s price range while city hotels came down from the upper price range, closing the gap between the two. How the summer decline through August looks against three years of monthly data is analyzed in detail in Tokyo August ADR: City Hotels −20.6%, Business −8.9%.

Seasonal shape, year over year — the spring peak and summer trough match last year; only the depth of the trough differs

Overlaying monthly estimated settled ADR year by year on a January–December axis separates differences in seasonal shape from differences in level. Business hotels in both 2025 and 2026 peak for the year in April (¥17,138 → ¥17,802), fall toward June, pass through a summer trough in August (¥12,689 → ¥11,554) and rise again from September. What differs is the depth of the trough: in 2026 the rate fell ¥6,248 from the April peak to the August trough, compared with a drop of only ¥4,449 in 2025. Having climbed higher than the prior year in spring, the summer fall looked that much larger.

Tokyo business hotels’ estimated settled ADR (tax-excl. equivalent). January–September are finalized; October–November are estimates based on current sales (dotted line, subject to change). Source: MetroEngines Research; compiled by HotelBank Editorial Team

City hotels share the same skeleton of a spring peak and summer trough, but in 2026 the rate fell ¥9,976 from ¥31,241 in April to ¥21,265 in August, a trough more than twice as deep as the prior year’s (April ¥31,076 → August ¥26,769, a drop of ¥4,307). January–March 2026 were all above the prior year, so rates held through winter into spring; city hotels’ year-on-year decline is better described not as “weak from the start of the year” but as a phase in which “the summer price range was cut from May onward.” September’s ¥24,650 recovered ¥3,385 from August, and the start of a move to win back the price range ahead of autumn demand is the starting point for thinking about rate design from next month on.

Tokyo city hotels’ estimated settled ADR (tax-excl. equivalent). January–September are finalized; October–November are estimates based on current sales (dotted line, subject to change). Source: MetroEngines Research; compiled by HotelBank Editorial Team

Note that October and November, shown as dotted lines on the charts, are current estimates (business October ¥20,204, N=893; November ¥19,007, N=897 / city October ¥35,652, N=108; November ¥33,573, N=108) and are calculated on a different basis from finalized months. A jump in level at the seam where finalized months give way to current estimates is a known behavior, and a simple comparison with the prior year’s finalized values (business October ¥15,348, November ¥16,540 / city October ¥31,970, November ¥32,756) must wait until those months are finalized at month-end. This article does not make October onward a subject of year-on-year comparison; it uses those months only as a directional reference that autumn price ranges are being set above summer’s.

Tokyo against six other major prefectures — the turn above last year is widespread, while the slow city-hotel recovery stands out in Tokyo

Viewed on its own, Tokyo’s September is “business hotels have turned, city hotels are halfway there,” but placing the same finalized September estimated settled ADR alongside six other major prefectures sharpens Tokyo’s position. For business hotels, Hokkaido +8.9%, Kanagawa +6.5% and Fukuoka +9.9% all came in above the prior year (Kyoto −2.4%), so Tokyo’s +2.2% is a modest gain among them. For city hotels, some regions far exceeded the prior year — Kanagawa +15.8%, Fukuoka +21.9% — while Tokyo, at −10.3%, did not even reach flat like Hokkaido (+0.1%). The breadth of the national business-hotel picture is laid out across all 47 prefectures in Business Hotel ADR Sep 2026: 44 Prefectures Up YoY, Median +7.9%.

Estimated settled ADR (tax-excl. equivalent; finalized month vs. finalized month). N= is the number of properties estimated. Osaka is compared against the 2025 Osaka-Kansai Expo period (April 13–October 13), and Aichi against a month including the Asian Games that opened on September 19, 2026; both involve special factors. Source: MetroEngines Research; compiled by HotelBank Editorial Team

Two of the comparison regions, however, carry special factors in either the prior or the current year. Osaka fell sharply, business −34.0% and city −21.8%, but September last year fell during the Osaka-Kansai Expo, so the comparison base itself is high. Aichi stands out at business +53.2% and city +53.7%, but includes demand from the Asian Games that opened on September 19, 2026. Excluding these two regions, the only places where city hotels ended September still below the prior year were Tokyo and Kyoto (−14.6%). That city-hotel recovery lags in both of these cities, each with a large volume of inbound overnight stays, is a useful clue when separating whether a Tokyo city hotel’s year-on-year gap is specific to the property or a market-wide issue.

September estimated OCC in the 90% range for both categories — occupancy held, rates made the difference

In September, when the year-on-year rate gap moved, what happened on the occupancy side? Monthly average estimated OCC (OTA listed-inventory basis) was 90.3% for business hotels (538–561 properties, 78,348–80,960 rooms, varying by day) and 90.0% for city hotels (52–54 properties, 16,054–17,294 rooms), almost level with each other. Going back through June–August, business hotels ran 90.9%, 91.7%, 88.8% and city hotels 89.8%, 89.5%, 87.2%, both staying around 90% apart from a slight easing in August, which shows the summer year-on-year decline was not accompanied by a collapse in occupancy. If rates alone fell below last year while occupancy held, the numbers suggest there may have been room to adjust on the pricing side.

Daily estimated OCC, September 2026 (OTA listed-inventory basis; based on observations from 45 days before the stay date to the latest). Business N=538–561 properties, city N=52–54 properties. Source: MetroEngines Research; compiled by HotelBank Editorial Team

By day, the month’s peak was concentrated in the first half of the 5-day holiday from Saturday, September 19 to Wednesday, September 23 (Autumnal Equinox Day): Sunday the 20th reached 98.0% for business and 96.5% for city, and Saturday the 19th 96.5% and 95.9%. The trough, conversely, came on the last day of the holiday, Wednesday the 23rd (Autumnal Equinox Day), the month’s lowest at 81.0% for business and 80.4% for city, and Thursday the 24th had not fully recovered at 83.5% and 84.5%. Stay dates falling on the last day of a holiday run see thin demand even when they are public holidays — a pattern that repeats every holiday period — and how to price the high night before a holiday against the low final day within the same holiday run determines revenue for the holiday week.

Averaged by day of week, business hotels are high on Saturday (91.6%) and Wednesday (91.3%) and low on Friday (88.5%) and Sunday (88.8%). City hotels, by contrast, peak sharply on Saturday (93.1%) and are low on Sunday (88.3%) and Friday (88.8%). Business hotels holding in the 90% range midweek (Tuesday–Thursday) reflects the resilience of business-trip and corporate demand, while city hotels forming a single peak on Saturday reflects their higher leisure share. Even with the same monthly average in the 90% range, business hotels show little day-of-week unevenness while city hotels concentrate on Saturday, and this difference in shape bears directly on how the weekly price range is set. The day-of-week shape on weekdays only, excluding the 5-day holiday, and how to translate it into inventory allocation, are explored in Tokyo Hotels Sep 2026 OCC: Business Peaks Wed 93.8%, City Wed & Sat.

Average of daily estimated OCC by day of week (OTA listed-inventory basis). Business N=538–561 properties, city N=52–54 properties. Source: MetroEngines Research; compiled by HotelBank Editorial Team

Mid-November booking curves — city hotels lead on Saturday, business hotels build more on weekdays

As a current read ahead of the autumn demand season, we checked the booking curves for Saturday, November 14 and Wednesday, November 18, using observations from 45 days out to the latest. Business hotels (542–544 properties, 79,073–79,096 rooms) built up 18.2 points for Saturday the 14th, from 64.8% at 45 days out to 76.9% at 30 days out to 83.0% at the latest (21 days out), and added 15.4 points for Wednesday the 18th, from 62.1% to 72.9% to 77.5% at the latest (25 days out). City hotels (52 properties, 16,029 rooms) went from 81.9% at 45 days out to 88.6% at 30 days out to 93.5% at the latest for Saturday the 14th, and from 78.7% to 86.5% to 88.8% for Wednesday the 18th — already around 80% at 45 days out, with only 10–12 points added from there.

Booking curves of estimated OCC (OTA listed-inventory basis), based on observations from 45 days before the stay date to the latest. Source: MetroEngines Research; compiled by HotelBank Editorial Team

What this shows is a difference in pacing: in autumn Tokyo, city hotels fill Saturdays early, while business hotels finish both weekdays and Saturdays by building from 30 days out to the last minute. Whether “the demand that builds close in was captured in the rate” is likely also part of why business hotels turned above last year in the finalized September figures, and the same pattern appears to be continuing toward November. Inventory sell-through moves daily, however, so the figures shown here are a snapshot as of October 25 and should be read on the understanding that they update as the stay date approaches. For the premise of reading estimated OCC curves without confusing them with sell-out rates, see also Estimated OCC vs Sell-Out Rate: Tokyo Business 77.7% vs Capsule 93.8%.

Revenue difference if the year-on-year gap narrows — three scenarios and rate × occupancy sensitivity

Taking the finalized September figures as the starting point, we estimate how much room revenue changes depending on where the year-on-year gap settles from October on. This is not a forecast of where results will land; it is a framework that applies past year-on-year rates to the prior-year month’s (September 2025) estimated settled ADR and compares the resulting levels on a “rate × occupancy” basis, multiplied by estimated OCC (September 2026 monthly average). Assuming 100 rooms and 30 nights, the gap for city hotels between the year-on-year decline staying at −10.3% and returning to flat is about ¥7.6 million a month — larger than the gap for business hotels between reverting to August’s year-on-year decline and returning to January–May growth (about ¥4.4 million).

Estimated settled ADR (tax-excl. equivalent) = finalized September 2025 value (business ¥13,072, city ¥27,467) × (1 + YoY rate). Rate × occupancy = estimated settled ADR × September 2026 monthly average estimated OCC (business 90.3%, city 90.0%), a trial figure used as a RevPAR guide. 100 rooms and 30 nights are assumptions. Source: MetroEngines Research; compiled by HotelBank Editorial Team

Whether moving rate or occupancy has more effect is easier to see on two axes. The two tables below show rate × occupancy when the finalized September rate is moved within ±10% and estimated OCC within 86–94% (the shaded cell is the base: finalized September rate at 90% occupancy). For business hotels, adding 5% to the rate (about ¥670) has almost the same effect as adding just over 4 points of occupancy, and the same relationship holds for city hotels. In Tokyo, where occupancy is already around 90%, there is likely more room to recover 5% on rate than to add a few more points of occupancy, which is consistent with this article’s observation that only rates fell below last year in the summer.

Rate × occupancy = estimated settled ADR (tax-excl. equivalent) × estimated OCC (OTA listed-inventory basis, assumed monthly average). A trial figure used as a RevPAR guide; it differs from actual room revenue and accounting figures. Source: MetroEngines Research; compiled by HotelBank Editorial Team

For revenue managers running business and city hotels in Tokyo — implications and action plan

Action plan based on this article’s data (no guarantee of results). Source: MetroEngines Research; compiled by HotelBank Editorial Team

Summary — three yardsticks for Tokyo’s summer into autumn

First, business hotels led the year-on-year turnaround. Finalized September estimated settled ADR for business hotels turned above the prior year for the first time in four months at ¥13,364 (+2.2%), while city hotels, at ¥24,650 (−10.3%), are still below last year but halved their decline from August’s −20.6%. Second, the year-on-year decline was a rate cut, not a collapse in occupancy. Estimated OCC for June–September held around 90% — 88.8–91.7% for business and 87.2–90.0% for city — and together with the compression of the city-to-business rate multiple from 2.10x to 1.84x, concessions at the upper price tier read as the main cause of the summer year-on-year gap. Third, autumn revenue will be decided by the precision of holiday and day-of-week pricing. In September, the first two days of the 5-day holiday (19th, 20th) were at 95.9–98.0% and the last day at 80.4–81.0%, a spread of 15–17 points within the same holiday, and by day of week business hotels showed a flat profile while city hotels concentrated on Saturday. Booking curves heading into November show city-hotel Saturdays reaching 80% at 45 days out and business hotels building 15–18 points from 45 days out to the latest; matching this pace pattern against your own property is likely to be the first step toward winning back the autumn price range.

Frequently Asked Questions

Q. Did estimated settled ADR for Tokyo business hotels exceed the prior year in September 2026?

A. Yes. Finalized September estimated settled ADR was ¥13,364 (N=914 properties), 2.2% above September 2025’s ¥13,072 (N=894 properties), turning around from the June–August year-on-year declines (−4.7%, −5.3%, −8.9%).

Q. Why are Tokyo city hotels still below the prior year?

A. Finalized September was ¥24,650, −10.3% year on year, so the decline continues, but it has halved from August’s −20.6%. Estimated OCC held at around 90%, and the core of the year-on-year gap was not a collapse in occupancy but a cut in rates at the upper price tier. Among the seven major prefectures, excluding Osaka and Aichi with their special factors, the only places where city hotels ended September still below the prior year were Tokyo and Kyoto.

Q. What does it mean when September estimated OCC is in the 90% range but rates are below the prior year?

A. It reads not as a drop in demand but as inventory being sold at lower rates than last year. Inventory may have been sold early at low prices, or price increases may not have kept up with demand building close in; the starting point is to overlay your property’s booking curve on the market curve.

Q. Can October and November 2026 estimated settled ADR be compared with the prior year?

A. This article does not compare them. October and November are estimates based on current sales, calculated on a different basis from finalized months, and the level appears to jump at the seam. Comparison with the prior year’s finalized values should wait until those months are finalized.

Q. How should the November 21–23 three-day weekend be priced?

A. Over September’s 5-day holiday, estimated OCC for the first two days (19th, 20th) reached 95.9–98.0%, while the last day, the 23rd, was the month’s lowest at 80.4–81.0%. Assuming the same pattern appears, there is room to design the ceiling for the first and middle days separately from the floor for the last day.

About the Data