MetroEngines Research estimated the September 2026 occupancy rate (estimated OCC) of city hotels in Japan from inventory listed on OTAs and other booking channels. Across the 21 prefectures for which the figure could be calculated — 425 properties and 77,583 rooms (based on major booking-site listings) — the median was 85.8%, and the room-weighted figure was 87.0%. By prefecture, Aomori (93.2%) and Kumamoto (93.0%) led the ranking; Tokyo, with 54 properties and 17,294 rooms, came in at 90.0%; Osaka was 82.0%; and Gunma, the lowest, was 73.2% — a spread of 20pt within the same segment. Meanwhile, the Japan Tourism Agency (JTA) “Overnight Travel Statistics Survey” first preliminary figures for September, published on October 30, put city hotel room occupancy at 74.6% (−1.0pt YoY). The gap with our estimate stems from differences in the denominator and the population measured. Continuing our series on occupancy by segment (business hotels, ryokan, and others), this article compares the estimated OCC of city hotels across 21 prefectures and lays out the gap versus business hotels in the same prefecture, the definitional differences from the JTA statistics, and where listed prices and estimated settled ADR sit — so that operators and revenue managers can check where their own prefecture stands.
• ADR (average daily rate) = an estimated settled rate (tax-excluded equivalent), calculated by applying a segment-specific adjustment factor to the lowest-priced plan each property publishes on OTAs (per room, two guests, tax included). Benchmarked against property-level results disclosed by listed hotel REITs (91 properties, most recent 3 months), the median error is about 7%. These are estimates and differ from each property’s actual settled rates and accounting figures.
• Area-level ADR (prefecture / municipality) is the median of the estimated settled rates of the properties in scope (the level of a typical property in the area).
• Wherever listed prices are mentioned, they refer to the per-room rate for two guests (tax included).
• OCC (occupancy rate) = rooms sold as a share of total rooms in the area (an estimate based on OTA sales inventory).
• Source: MetroEngines Research (continuously tracks OTA published prices for approximately 168,000 properties in Japan, of which approximately 27,000 are confirmed active)
Key Takeaways
— Median 85.8%, weighted 87.0%: the September 2026 estimated city hotel OCC across the 21 prefectures where it could be calculated (425 properties, 77,583 rooms). The range runs from Aomori at 93.2% to Gunma at 73.2%; the remaining 26 prefectures lacked sufficient inventory data and were not calculated.
— Saturday 90.8%, Mon–Thu 85.8%: simple averages across the 21 prefectures. During the five-day holiday from September 19, listed inventory in Osaka, Kyoto and Fukuoka was almost fully absorbed, while the stay night of September 23 (Wed), the last day of the holiday, was the month’s low in many prefectures.
— Business hotels ahead in 13 prefectures, city hotels ahead in 8: in Kagoshima, city hotels were at 81.1% vs 63.8% for business hotels (+17.3pt), while Gunma was the reverse at −10.2pt. In Hokkaido, Aichi and Aomori, business hotels were ahead.
— City hotels at 74.6% in JTA statistics: the official statistics use all rooms as the denominator, whereas our estimate measures absorption of inventory listed on OTAs and other channels — they measure different things. The gap should be read as a definitional difference; we do not argue which is “correct.”
Three Layers of “City Hotel” Definitions — The Population This Article Counts
The first point to note is that the “city hotel” category differs across three layers: law, statistics, and OTAs. Japan’s Hotel Business Act has no “city hotel” category; under licensing, everything from ryokan to luxury hotels falls under the same “ryokan/hotel business.” The JTA’s Overnight Travel Statistics Survey defines it as one facility type — “properties located in urban areas other than resort hotels and business hotels” — and leaves the choice of type to each property’s self-reporting. Booking sites classify properties by their own listing categories, tied to neither the law nor the statistics. We covered these three layers in detail in “What Is a City Hotel? 3 Definitions”, so here we carry over only the conclusion.
In this article, “city hotel” refers to MetroEngines Research’s segment classification, which consolidates the listing categories of major booking sites at the property level. MetroEngines Research tracks approximately 168,000 properties in Japan and analyzes the approximately 27,000 properties (about 1.26 million rooms) confirmed as active on OTAs. The estimated-OCC scope within that is a fixed cohort of properties that list at least 30% of their total rooms on OTAs and other channels and that could be observed as the same set of properties on every stay date in September 2026; property and room counts by prefecture are shown in the tables below. In other words, the figures here describe absorption not at “all city hotels in the prefecture” but at “city hotels that put a certain share of inventory on major booking sites,” and this population definition is the main driver of the gap with the JTA statistics. The occupancy rate is defined in the standard way: rooms sold divided by rooms available.
Estimated OCC Ranking Across 21 Prefectures — From Aomori 93.2% to Gunma 73.2%
Now to the main topic: levels by prefecture. Of the 47 prefectures, the September 2026 estimated OCC could be calculated for 21. The top spots went to prefectures with around 10 properties each: Aomori at 93.2% (9 properties, 1,020 rooms), Kumamoto at 93.0% (13 properties, 1,481 rooms) and Nagano at 91.1% (15 properties, 1,446 rooms). Among prefectures with larger samples, Hokkaido (90.6%, 35 properties, 8,550 rooms) and Tokyo (90.0%, 54 properties, 17,294 rooms) reached the 90% range, followed by Aichi at 88.4% (22 properties, 3,640 rooms) and Fukuoka at 87.1% (22 properties, 2,949 rooms). Kyoto came in at 85.3% (36 properties, 5,050 rooms) and Osaka at 82.0% (34 properties, 7,905 rooms), placing Kansai’s two major cities in the middle to lower part of the ranking.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data
Note that in prefectures with fewer than 10 properties (Aomori, Shimane, Gunma), inventory moves by a single property can easily shift the prefecture’s figure. In addition, four prefectures were observed on fewer than 30 days. Shimane was observed on only 10 days in September and should be treated as a reference value; Aomori (24 days), Miyagi (26 days) and Gunma (26 days) also have monthly averages that do not represent every day. The table shows days observed alongside Saturday and Monday–Thursday averages.
| Rank | Prefecture | Est. OCC (Sep avg.) | Properties | Rooms | Days observed | Saturday avg. | Mon–Thu avg. |
|---|---|---|---|---|---|---|---|
| 1 | Aomori ※ | 93.2% | 9 properties | 1,020 rooms | 24 days | 95.9% | 93.5% |
| 2 | Kumamoto | 93.0% | 13 properties | 1,481 rooms | 30 days | 93.1% | 93.5% |
| 3 | Nagano | 91.1% | 15 properties | 1,446 rooms | 30 days | 95.5% | 91.4% |
| 4 | Hokkaido | 90.6% | 35 properties | 8,550 rooms | 30 days | 91.7% | 91.4% |
| 5 | Miyagi ※ | 90.3% | 10 properties | 1,047 rooms | 26 days | 98.8% | 89.6% |
| 6 | Tokyo | 90.0% | 54 properties | 17,294 rooms | 30 days | 93.1% | 89.9% |
| 7 | Hiroshima | 88.6% | 14 properties | 1,662 rooms | 30 days | 89.1% | 90.1% |
| 8 | Aichi | 88.4% | 22 properties | 3,640 rooms | 30 days | 91.7% | 88.1% |
| 9 | Fukuoka | 87.1% | 22 properties | 2,949 rooms | 30 days | 92.8% | 85.6% |
| 10 | Chiba | 86.0% | 24 properties | 7,297 rooms | 30 days | 91.4% | 85.1% |
| 11 | Hyogo | 85.8% | 23 properties | 3,446 rooms | 30 days | 90.7% | 85.2% |
| 12 | Ehime | 85.6% | 10 properties | 1,304 rooms | 30 days | 87.2% | 86.5% |
| 13 | Kanagawa | 85.5% | 28 properties | 5,477 rooms | 30 days | 91.9% | 83.8% |
| 14 | Kyoto | 85.3% | 36 properties | 5,050 rooms | 30 days | 89.7% | 84.5% |
| 15 | Shimane ※ | 83.2% | 9 properties | 688 rooms | 10 days | 93.5% | 82.3% |
| 16 | Shizuoka | 83.2% | 19 properties | 2,011 rooms | 30 days | 87.0% | 83.7% |
| 17 | Osaka | 82.0% | 34 properties | 7,905 rooms | 30 days | 88.9% | 80.4% |
| 18 | Nagasaki | 81.8% | 15 properties | 1,483 rooms | 30 days | 89.4% | 80.1% |
| 19 | Kagoshima | 81.1% | 14 properties | 1,844 rooms | 30 days | 86.8% | 80.6% |
| 20 | Ibaraki | 79.7% | 10 properties | 1,092 rooms | 30 days | 83.6% | 83.6% |
| 21 | Gunma ※ | 73.2% | 9 properties | 897 rooms | 26 days | 85.5% | 73.0% |
| — | 21 prefectures: total / median | 85.8% (room-weighted 87.0%) | 425 properties | 77,583 rooms | — | 90.8% | 85.8% |
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data (※ = observed on fewer than 30 days. Estimated OCC is the average of daily values in September; Saturday and Mon–Thu figures are averages of daily values on those days of the week)
The remaining 26 prefectures (Mie, Saga, Wakayama, Saitama, Oita, Nara, Miyazaki, Toyama, Yamaguchi, Yamagata, Yamanashi, Gifu, Okayama, Iwate, Tokushima, Niigata, Tochigi, Okinawa, Shiga, Ishikawa, Fukui, Fukushima, Akita, Kagawa, Kochi and Tottori) did not have enough city hotels listing at least 30% of their rooms on major booking sites to form a fixed cohort, so September estimated OCC was not calculated. We list these explicitly as “prefectures that could not be calculated” and do not produce estimates for them. Large hotel markets such as Okinawa and Ishikawa are included because a high share of their hotels are listed as resort or business hotels — not because city hotels do not exist there.
Day of Week and the Five-Day Holiday — Saturday 90.8%, the Holiday’s Last Night (Wed, Sep 23) the Month’s Low
Next, we look at movement within the month. On a simple average across the 21 prefectures, estimated OCC for Saturday stays was 90.8% versus 85.8% for Monday–Thursday, a day-of-week gap of only around 5pt. In prefectures with strong business-travel demand, weekdays sometimes exceeded Saturdays: Hiroshima recorded 90.1% Mon–Thu vs 89.1% on Saturday, and Kumamoto 93.5% Mon–Thu vs 93.1% on Saturday, pointing to a weekday-driven demand structure. By contrast, Ibaraki showed 83.6% on Saturday but an average of 63.6% on Sunday, with the lowest Sunday dropping to 50.8% — a swing that signals a strong weekend-leisure profile.
The month’s peak came during the five-day holiday from Saturday, September 19 to Wednesday, September 23. For the stay night of Sunday, September 20, estimated OCC in Osaka, Kyoto, Fukuoka and Aichi reached the 99% range to 100%, with listed inventory on major booking sites almost fully absorbed. Tokyo also held at 95–96% from Saturday, September 19 through Monday, September 21. However, for the stay night of Wednesday, September 23 — the holiday’s last day — 8 of the 21 prefectures recorded their monthly low, including Tokyo at 80.4% and Hokkaido at 77.9%, and Osaka also fell to 74.6%. Demand is thin on the final night of a holiday when the next day is a workday. This “trough on the eve of the return to work” is a starting point for designing multi-night plans and pricing for the final night. Our Silver Week 2026 actuals, which reviewed the entire holiday with daily results across 7 areas × 4 segments, shows the same shape: a peak on the night of September 20 and a trough on the final night of September 23.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data (Tokyo 54 properties / 17,294 rooms; Osaka 34 properties / 7,905 rooms; Hokkaido 35 properties / 8,550 rooms; based on major booking-site listings)
Gap Versus Business Hotels in the Same Prefecture — Kagoshima +17.3pt, Gunma −10.2pt
City hotel levels become more meaningful when set against business hotels in the same prefecture. Calculating business hotel estimated OCC for the same month with the same method, business hotels exceeded city hotels in 13 of the 21 prefectures, and city hotels were ahead in 8. The largest gap was in Kagoshima, where city hotels were at 81.1% (14 properties) versus 63.8% for business hotels (72 properties), putting city hotels 17.3pt ahead. Ehime also showed city hotels at 85.6% vs 79.7% for business hotels (+5.9pt), indicating that city hotels filled first in Shikoku and southern Kyushu. With Osaka at +2.0pt, Nagasaki +2.9pt and Ibaraki +2.8pt, there are eight prefectures where city hotels lead.
Conversely, Hokkaido (city 90.6% vs business 92.8%, −2.2pt), Aichi (88.4% vs 91.4%, −3.0pt) and Aomori (93.2% vs 96.9%, −3.7pt) saw business hotels fill first. Gunma had the largest gap at −10.2pt, but given that its city hotel cohort is just 9 properties observed on 26 days, the sample is too thin to call this a prefecture-wide structure. In some prefectures the segment gap barely registers: Tokyo was 90.0% vs 90.3% (−0.3pt) and Nagano −0.1pt. How estimated settled ADR moved in opposite directions for Hokkaido’s city and business hotels over the summer is covered in our segment analysis of Hokkaido city and business hotels; reading occupancy gaps together with rate gaps provides material for autumn and winter pricing.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data
| Prefecture | City hotel est. OCC | Business hotel est. OCC | Gap (city − business) |
|---|---|---|---|
| Kagoshima | 81.1% (14 properties) | 63.8% (72 properties) | +17.3pt |
| Ehime | 85.6% (10 properties) | 79.7% (50 properties) | +5.9pt |
| Nagasaki | 81.8% (15 properties) | 78.9% (57 properties) | +2.9pt |
| Ibaraki | 79.7% (10 properties) | 76.9% (78 properties) | +2.8pt |
| Osaka | 82.0% (34 properties) | 80.0% (285 properties) | +2.0pt |
| Kumamoto | 93.0% (13 properties) | 91.8% (58 properties) | +1.2pt |
| Chiba | 86.0% (24 properties) | 85.1% (85 properties) | +0.9pt |
| Kyoto | 85.3% (36 properties) | 84.8% (190 properties) | +0.5pt |
| Nagano | 91.1% (15 properties) | 91.2% (75 properties) | -0.1pt |
| Tokyo | 90.0% (54 properties) | 90.3% (539 properties) | -0.3pt |
| Hiroshima | 88.6% (14 properties) | 89.5% (93 properties) | -0.9pt |
| Miyagi | 90.3% (10 properties) | 91.6% (58 properties) | -1.3pt |
| Kanagawa | 85.5% (28 properties) | 87.0% (116 properties) | -1.5pt |
| Shimane | 83.2% (9 properties) | 85.1% (30 properties) | -1.9pt |
| Hokkaido | 90.6% (35 properties) | 92.8% (222 properties) | -2.2pt |
| Hyogo | 85.8% (23 properties) | 88.0% (74 properties) | -2.2pt |
| Shizuoka | 83.2% (19 properties) | 85.5% (117 properties) | -2.3pt |
| Fukuoka | 87.1% (22 properties) | 90.0% (175 properties) | -2.9pt |
| Aichi | 88.4% (22 properties) | 91.4% (172 properties) | -3.0pt |
| Aomori | 93.2% (9 properties) | 96.9% (35 properties) | -3.7pt |
| Gunma | 73.2% (9 properties) | 83.4% (48 properties) | -10.2pt |
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data (September 2026; both segments are fixed cohorts of properties listing at least 30% of total rooms on major booking sites)
How Bookings Build — Tokyo and Osaka City Hotels Reach the 60% Range at LT45, Ahead of Business Hotels
For revenue managers, it matters not only where occupancy lands but how full a hotel was at each point along the way. We compared booking curves — averaging all September stay dates by lead time (LT, days before check-in) — for city and business hotels in Tokyo and Osaka. Tokyo city hotels (54 properties / 16,971 rooms, monthly average, based on major booking-site listings) had reached an estimated OCC of 64.8% at LT45, building to 71.3% at LT30, 80.0% at LT14, 83.4% at LT7 and 90.0% the day before (LT1). Tokyo business hotels (550 properties / 79,533 rooms) were at 51.9% at LT45, 72.9% at LT14 and 90.3% at LT1. While both landed at almost the same level, city hotels were ahead by 13pt at LT45 and 7pt at LT14.
Osaka followed the same pattern: city hotels (36 properties / 8,192 rooms) were at 63.4% at LT45, 72.0% at LT14 and 82.0% at LT1, while business hotels (294 properties / 48,614 rooms) were at 45.0% at LT45, 60.9% at LT14 and 80.0% at LT1. The shape of the curves likely reflects a segment difference: city hotel rooms are built up early through group, banquet and inbound bookings, while business hotels fill last with short-lead business travel. The curves in this article use LT45 through LT1; observation points earlier than 45 days are excluded because they include periods before listing allotments are loaded. Our booking curve comparison of city and business hotels in Miyagi, which tracks the gap from the same LT45 starting point, is also a useful reference for reading segment differences.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data (average of all September 2026 stay dates. Tokyo city 54 properties / 16,971 rooms, Tokyo business 550 properties / 79,533 rooms, Osaka city 36 properties / 8,192 rooms, Osaka business 294 properties / 48,614 rooms; based on major booking-site listings)
The Gap With the JTA’s 74.6% for City Hotels — Different Denominators, Different Populations
The estimated OCC figures so far are mostly in the 80–90% range, but in the JTA’s Overnight Travel Statistics Survey first preliminary figures for September 2026, city hotel room occupancy was 74.6% (−1.0pt YoY), with the overall figure at 62.6%, business hotels 76.9%, resort hotels 60.5% and ryokan 43.8%. It is not appropriate to debate the gap in terms of “which is correct,” because the two measure different things. See also our article on reading Okinawa occupancy: JTA figures vs OTA inventory-based estimates, which lays out the same definitional difference as a 27.5pt gap in Okinawa.
First, the denominators differ. JTA room occupancy is rooms used as a share of all rooms at surveyed properties, and the first preliminary figures are estimated from questionnaires with an effective response rate of 30.8%. Our estimated OCC uses a fixed cohort of properties listing at least 30% of total rooms on OTAs and other channels, and divides total rooms minus remaining listed rooms by total rooms. Rooms that are not listed (sold through the hotel’s own website, corporate contracts, groups, stays bundled with banquets, etc.) are counted on the absorbed side, so the figure is structurally higher. Second, the populations differ. The JTA estimates properties nationwide through a full survey of properties with 20 or more rooms and a sample survey of those with 1–19 rooms, with facility type self-reported. We classify properties by the listing categories of major booking sites and count only those meeting the listing-share condition. Third, the latest published prefecture-level figures are the August second preliminary figures, a month off from this article’s September data. Note also that the change in stratification criteria from the January 2026 survey onward affects year-on-year comparisons in the official statistics.
For reference, we placed the JTA’s August second preliminary city hotel room occupancy by prefecture alongside our September estimated OCC. Hokkaido was 83.1% (JTA) vs 90.6% (ours), Tokyo 70.6% vs 90.0%, and Osaka 71.2% vs 82.0%; in many prefectures the gap falls within roughly 5–20pt. The largest gap was in Kumamoto, at 49.1% (JTA) vs 93.0% (ours), a 43.9pt difference. After the Kumamoto earthquake of July 28, Kumamoto City began on August 3 a support program offering accommodation facilities as secondary evacuation shelters; in situations like this, where rooms are used outside major booking sites, the gap between our estimate — which counts unlisted rooms as absorbed — and official statistics that use all rooms as the denominator tends to widen. Kagoshima (58.0% vs 81.1%) and Ehime (64.6% vs 85.6%) also showed gaps of more than 20pt, and in both prefectures city hotels also led business hotels in the same-prefecture comparison. These are prefectures where absorption of listed inventory and all-room occupancy move differently, so both indicators need to be watched.
| Prefecture | MetroEngines est. OCC Sep 2026, OTA-listed inventory basis | JTA city hotel room occupancy Aug 2026, 2nd preliminary | Gap (reference only; definitions and months differ) |
|---|---|---|---|
| Aomori | 93.2% | 79.8% | +13.4pt |
| Kumamoto | 93.0% | 49.1% | +43.9pt |
| Nagano | 91.1% | 84.9% | +6.2pt |
| Hokkaido | 90.6% | 83.1% | +7.5pt |
| Miyagi | 90.3% | 76.7% | +13.6pt |
| Tokyo | 90.0% | 70.6% | +19.4pt |
| Hiroshima | 88.6% | 77.6% | +11.0pt |
| Aichi | 88.4% | 71.9% | +16.5pt |
| Fukuoka | 87.1% | 78.6% | +8.5pt |
| Chiba | 86.0% | 74.7% | +11.3pt |
| Hyogo | 85.8% | 75.6% | +10.2pt |
| Ehime | 85.6% | 64.6% | +21.0pt |
| Kanagawa | 85.5% | 68.4% | +17.1pt |
| Kyoto | 85.3% | 70.1% | +15.2pt |
| Shimane | 83.2% | 78.4% | +4.8pt |
| Shizuoka | 83.2% | 73.3% | +9.9pt |
| Osaka | 82.0% | 71.2% | +10.8pt |
| Nagasaki | 81.8% | 75.2% | +6.6pt |
| Kagoshima | 81.1% | 58.0% | +23.1pt |
| Ibaraki | 79.7% | 68.8% | +10.9pt |
| Gunma | 73.2% | 57.1% | +16.1pt |
Source: Compiled by HotelBank Editorial Team from Japan Tourism Agency “Overnight Travel Statistics Survey” (September 2026 1st preliminary and August 2026 2nd preliminary, published October 30, 2026) and MetroEngines Research data. Gaps are for reference only; definitions, populations and months differ
In the JTA’s long-term series, city hotel room occupancy fell from 79.5% in 2019 to 34.1% in 2020, then recovered to 68.8% in 2023, 72.3% in 2024 and 74.1% in 2025. September 2026’s 74.6% is slightly above the 2025 full-year level; even with a −1.0pt YoY change, the level itself remains on the recovery trajectory. Monthly trends in 2026 for ryokan and the overall market are covered in our article on August ryokan occupancy of 47.4%.
Price Levels — Listed Prices Alongside Estimated Settled ADR (Supplementary Indicator)
Occupancy levels can mislead unless read together with rates. We therefore set the September listed prices (all-plan average, two guests per room, tax included) of the estimated-OCC cohort alongside the estimated settled ADR of city hotels in each prefecture (an estimate applying a segment-specific factor to the lowest-plan level, tax-excluded equivalent, median). Estimated settled ADR is the official ADR indicator used in our articles since July 9, 2026; because it differs from listed prices in both level and definition, the two are not added or subtracted even when shown in the same table.
The highest estimated settled ADR levels were in Tokyo (¥25k, N=110), Kyoto (¥20k, N=61), Fukuoka (¥19k, N=43) and Aichi (¥18k, N=31), while Ehime (¥8k, N=23), Kagoshima (¥8k, N=26), Ibaraki (¥8k, N=17), Shimane (¥9k, N=10) and Gunma (¥9k, N=13) were below ¥10k. Year on year, Aichi stood out at +53.7%, likely driven largely by the overlap with the Asian Games, which opened on September 19. Fukuoka (+21.9%), Hiroshima (+18.9%), Miyagi (+28.9%) and Aomori (+49.5%) were also above the prior year. The details of the finalized September figures for Fukuoka and Hiroshima are broken down by segment in Fukuoka’s finalized September estimated settled ADR and Hiroshima’s finalized September.
On the other hand, Tokyo (−10.3%), Osaka (−21.8%) and Kyoto (−14.6%) were below the prior year. September 2025 fell during Expo 2025 Osaka, Kansai, when Kansai rates were elevated, and the rebound effect is likely showing up in Osaka and Kyoto. In terms of estimated OCC, Osaka (82.0%) and Kyoto (85.3%) sit in the middle or lower part of the 21 prefectures, so the YoY rate decline needs to be checked together with the occupancy level. High occupancy and high rates are different things: a prefecture like Aomori, with estimated OCC of 93.2% and estimated settled ADR of ¥13k, should not be treated as the same kind of “high occupancy” as Tokyo at 90.0% and ¥25k.
| Prefecture | Listed price Sep, avg. of est.-OCC cohort, tax incl. | Est. settled ADR Sep 2026 Median of prefecture city hotels, tax-excl. equiv. | Est. settled ADR Sep 2025 | YoY |
|---|---|---|---|---|
| Tokyo | ¥38,906 | ¥24,650 (N=110) | ¥27,467 (N=107) | -10.3% |
| Kyoto | ¥35,079 | ¥20,179 (N=61) | ¥23,624 (N=58) | -14.6% |
| Fukuoka | ¥29,020 | ¥18,547 (N=43) | ¥15,212 (N=37) | +21.9% |
| Aichi | ¥47,844 | ¥18,141 (N=31) | ¥11,803 (N=30) | +53.7% |
| Hokkaido | ¥28,541 | ¥16,485 (N=75) | ¥16,476 (N=74) | +0.1% |
| Kanagawa | ¥26,299 | ¥14,932 (N=42) | ¥12,890 (N=40) | +15.8% |
| Shizuoka | ¥24,036 | ¥14,565 (N=31) | ¥13,506 (N=31) | +7.8% |
| Aomori | ¥24,222 | ¥13,128 (N=16) | ¥8,781 (N=16) | +49.5% |
| Osaka | ¥25,162 | ¥13,077 (N=97) | ¥16,733 (N=91) | -21.8% |
| Hiroshima | ¥26,899 | ¥12,897 (N=31) | ¥10,849 (N=29) | +18.9% |
| Hyogo | ¥22,556 | ¥12,737 (N=47) | ¥13,808 (N=46) | -7.8% |
| Chiba | ¥20,625 | ¥12,041 (N=53) | ¥11,017 (N=51) | +9.3% |
| Nagano | ¥21,703 | ¥11,638 (N=16) | ¥11,081 (N=16) | +5.0% |
| Miyagi | ¥24,600 | ¥11,236 (N=21) | ¥8,720 (N=20) | +28.9% |
| Nagasaki | ¥23,881 | ¥11,173 (N=25) | ¥11,186 (N=25) | -0.1% |
| Kumamoto | ¥25,789 | ¥10,870 (N=19) | ¥9,702 (N=19) | +12.0% |
| Gunma | ¥16,999 | ¥9,071 (N=13) | ¥10,249 (N=11) | -11.5% |
| Shimane | ¥15,078 | ¥8,508 (N=10) | ¥8,776 (N=10) | -3.1% |
| Ibaraki | ¥16,674 | ¥8,431 (N=17) | ¥8,808 (N=18) | -4.3% |
| Kagoshima | ¥17,435 | ¥8,292 (N=26) | ¥9,656 (N=23) | -14.1% |
| Ehime | ¥11,628 | ¥7,737 (N=23) | ¥7,876 (N=21) | -1.8% |
Source: Compiled by HotelBank Editorial Team from MetroEngines Research data (listed prices are the monthly average of September daily averages for the estimated-OCC cohort, tax included. Estimated settled ADR is the median of city hotels in each prefecture, tax-excluded equivalent; N is the number of properties calculated. YoY refers to estimated settled ADR)
Implications for Operators and Revenue Managers — How to Use Your Prefecture’s Position
When applying these figures to your own decisions, we suggest three steps. First, confirm which segment category your property is listed under on major booking sites, and before comparing the city hotel column here with your actual occupancy, check that it matches the facility type you report in the JTA statistics. If the classification is off, so is the benchmark. Second, look at the direction of the gap versus business hotels in the same prefecture. In prefectures where city hotels lead, such as Kagoshima, Ehime and Osaka, there is room to convert early holiday and weekend bookings into higher rates; in prefectures where business hotels lead, such as Hokkaido, Aichi and Aomori, capturing weekday corporate and group demand is the upside for city hotels. Third, look at the shape of the booking curve. Tokyo and Osaka city hotels build to the 60% range by LT45, so price steps between LT30 and LT14 are the core revenue opportunity.
The fact that the stay night of Wednesday, September 23 — the holiday’s last day — was the month’s low in many prefectures also has implications for multi-night plan design. Capturing high single-night rates on the high-demand days early in a holiday while attaching perks to multi-night stays that include the final night could help lift occupancy on the trough day. Even in Tokyo, Osaka and Kyoto, where estimated settled ADR was below the prior year, estimated OCC was in the 80–90% range, meaning listed inventory was still absorbed at a certain level even as rates ran below last year. Heading into the year-end season, a realistic next step is to check September’s absorption patterns by day of week and lead time in your prefecture’s column and to make weekend and holiday price steps one notch finer.
Summary
City hotel estimated OCC in September 2026 was a median of 85.8% and a room-weighted 87.0% across the 21 prefectures where it could be calculated (425 properties, 77,583 rooms). Aomori (93.2%) and Kumamoto (93.0%) led, with Tokyo at 90.0%, Osaka 82.0% and Gunma 73.2% — a 20pt spread within the same segment. Compared with business hotels in the same prefecture, business hotels were ahead in 13 prefectures, while Kagoshima showed the largest gap with city hotels ahead by +17.3pt. The JTA’s Overnight Travel Statistics Survey first preliminary figure for September was 74.6% for city hotels; the gap with our estimate reflects differences in the denominator (all rooms vs listed inventory) and population (full-population estimate vs fixed cohort), not a question of which is correct. On rates, Aichi, Fukuoka and Hiroshima were above the prior year in estimated settled ADR, while Tokyo, Osaka and Kyoto were below. The remaining 26 prefectures were not calculated due to insufficient inventory data, so please do not treat these figures as national values. Data as of November 4, 2026.
Frequently Asked Questions
Q. What was the city hotel occupancy rate in September 2026?
A. The occupancy rate estimated by MetroEngines Research from inventory listed on OTAs and other channels (estimated OCC) was a median of 85.8% across the 21 prefectures where it could be calculated (425 properties, 77,583 rooms), or 87.0% weighted by rooms. The highest was Aomori at 93.2% and the lowest Gunma at 73.2%, with Tokyo at 90.0%. In the JTA’s Overnight Travel Statistics Survey first preliminary figures for September 2026, city hotel room occupancy was 74.6%; the two differ in definition and population.
Q. Why are estimated OCC and the JTA’s room occupancy at different levels?
A. JTA room occupancy is rooms used as a share of all rooms at surveyed properties, aggregated by the facility type each property reports. Our estimated OCC uses a fixed cohort of properties listing at least 30% of total rooms on OTAs and other channels and is calculated by subtracting remaining listed rooms from total rooms, so unlisted rooms are counted as absorbed and the figure comes out higher. They are not a matter of which is correct; they are indicators that measure different things.
Q. Which has higher occupancy, city hotels or business hotels?
A. On September 2026 estimated OCC, business hotels exceeded city hotels in 13 of 21 prefectures, and city hotels were ahead in 8. The largest gap was in Kagoshima, where city hotels were at 81.1% vs 63.8% for business hotels (+17.3pt). Nationally, the JTA’s September first preliminary figures show business hotels at 76.9% and city hotels at 74.6%, with business hotels 2.3pt higher.
Related Articles
References and Sources
- Japan Tourism Agency, “Overnight Travel Statistics Survey (September 2026 1st preliminary, August 2026 2nd preliminary),” press release (October 30, 2026) (Japanese)
- Japan Tourism Agency, “Overnight Travel Statistics Survey, September 2026 (1st preliminary) results,” Table 5: Room occupancy by facility type (Japanese)
- Japan Tourism Agency, “Overnight Travel Statistics Survey” statistics page (survey overview and completion guidelines) (Japanese)
- Kumamoto City, “2026 Kumamoto Earthquake: Secondary Evacuation to Accommodation Facilities (Accommodation Support Program)” (August 2026) (Japanese)
- Aichi Now (Aichi Prefecture official tourism site), “The 20th Asian Games (September 19 – October 4, 2026)”
- MetroEngines Research — estimated OCC and booking curves by prefecture × segment (September 2026 stays; fixed cohort of properties listing at least 30% of total rooms on major booking sites), estimated settled ADR and listed prices by prefecture × segment (retrieved November 4, 2026)