Search for “Okinawa occupancy rate” and two numbers come up. The Japan Tourism Agency’s Overnight Travel Statistics Survey puts Okinawa Prefecture’s room occupancy at 61.8% for July 2026 (second preliminary figures), while MetroEngines Research’s OTA listed-inventory-based occupancy for Okinawa, estimated from rooms listed on OTAs and similar channels, was 89.3% in the same July and reached 91.5% for October 1–19. Neither figure is wrong; they measure different things. Written for operators and investors in Okinawa hotels and for those considering entering from outside the prefecture, this article sets the two occupancy rates side by side by definition, explains why they differ by 27.5pt in the same month, and sorts out which one to use for which decision.
• ADR (average daily rate) = an estimated settled rate (tax-excluded equivalent), calculated by applying a segment-specific adjustment factor to the lowest plan each property publishes on OTAs (two guests per room, per-room rate, tax included). Checked 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 transaction prices and accounting figures.
• Area-level (prefecture / municipality) ADR is the median of the estimated settled rates of the properties covered (the level of a typical property in the area).
• Where listed prices are cited, they are per-room rates 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 published OTA prices for about 168,000 properties in Japan, of which about 27,000 are confirmed as operating)
• Japan Tourism Agency room occupancy rate (official figure)= the figure the Japan Tourism Agency publishes monthly by prefecture and property type in its Overnight Travel Statistics Survey. It is an actual result based on responses from surveyed properties and counts rooms used regardless of sales channel. First preliminary figures are later revised in the second preliminary and final figures.
• OTA listed-inventory-based occupancy (our estimate)= 100 − 100 × rooms remaining on OTAs ÷ total rooms. It covers a fixed cohort of properties listing at least 30% of their total rooms on OTAs and similar channels (933 properties / 37,131 rooms in Okinawa in October 2026, the same set for every stay date). Because rooms not listed on OTAs are counted as already sold, it runs higher than actual occupancy. Only observations within 45 days before the stay date are used.
- — 61.8% and 89.3%: In July 2026, Okinawa’s Japan Tourism Agency room occupancy (second preliminary) was 61.8% and its OTA listed-inventory-based occupancy (our estimate) was 89.3%, a gap of 27.5pt.
- — The gap comes from the definitions: The main causes are counting rooms not listed on OTAs as sold, and the different set of properties covered (most ryokan and simple lodgings are excluded). The gap was 26.7–29.8pt in May–July 2026.
- — The gap differs by segment: In July 2026 the gap was 11.2pt for business hotels and 20.8pt for resort hotels. Actual resort-hotel occupancy was only 69.5% on the Japan Tourism Agency figure.
- — October 1–19: 91.5%: The peak was 96.1% on Saturday, October 10, the first day of a three-day weekend; the low was 87.2% on Thursday, October 1. The daily range narrowed from 15.4pt in September to 8.9pt.
- — Which to use: Use the Japan Tourism Agency figure for full-year actual occupancy and business-plan assumptions, and the OTA inventory-based figure for near-term demand strength and day-of-week and segment differences. Estimated settled ADR was above the same month of the previous year in every month from July to October 2026 (+2.7% to +18.2%).
Okinawa has two occupancy rates — official figures vs. inventory-based estimates: get the definitions straight first
Let us first lay out the differences in definition. The Japan Tourism Agency’s room occupancy rate is an actual result calculated from the rooms used and total rooms that properties report on the survey form, and it counts “filled rooms” including sales that do not go through OTAs, such as direct bookings, travel agencies, corporate contracts and groups. Its population is broad, covering ryokan, resort hotels, business hotels, city hotels, simple lodgings and company/group lodging facilities. OTA listed-inventory-based occupancy, by contrast, is an estimate that gauges demand strength from changes in rooms remaining on OTAs. It is timely, but structurally it cannot distinguish rooms that are not put on OTAs at all.
| Item | Japan Tourism Agency room occupancy (official) | OTA listed-inventory-based occupancy (our estimate) |
|---|---|---|
| Calculation | Actual result: rooms used, as reported by surveyed properties, divided by rooms available (operating-day basis) | 100 − 100 × rooms remaining on OTAs ÷ total rooms (daily observation, lead time within 45 days) |
| Properties covered | Lodging facilities in Okinawa Prefecture (ryokan, resort, business and city hotels, simple lodgings, company/group lodging) | Fixed cohort listing at least 30% of total rooms on OTAs and similar channels (October 2026: 933 properties / 37,131 rooms) |
| Sales outside OTAs | Included (direct, travel agencies, corporate contracts, groups) | Cannot be distinguished; unlisted rooms are counted as sold → runs higher than actual occupancy |
| Release timing | First preliminary (national) at the end of the following month; prefecture-level one month later, at the end of the month after next. Revised in the second preliminary and final figures | Updated daily; the current month can be viewed as an interim reading |
| Best suited for | Full-year actual occupancy, assumptions for business plans and loan screening, year-on-year trends | Demand strength now and a few weeks out, differences by day of week, holiday and segment, pricing decisions |
Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey and MetroEngines Research
How to read the statistics themselves is covered in What Is Japan’s Accommodation Survey? First Estimates Move 0.6pt and Overnight Travel Statistics Survey Guide, so this article focuses on Okinawa’s numbers.
Official figures — Okinawa’s room occupancy was 61.8% in July 2026, 2.6pt below the same month a year earlier
In the Japan Tourism Agency’s prefecture-level monthly room occupancy from the Overnight Travel Statistics Survey, the latest figure for Okinawa is 61.8% for July 2026 (second preliminary). The August 2026 figures released on September 30, 2026 are first preliminary figures and give only the national value (64.4%); because prefecture-level figures are released one month later, there is no August figure for Okinawa yet. All 2026 (Reiwa 8) values in the time-series table are preliminary and should be read on the assumption that they will be revised.
From January to July 2026 Okinawa moved through 51.4%, 59.7%, 58.3%, 59.9%, 56.7%, 55.1% and 61.8%. Against the same months a year earlier that is −1.1pt in January, −2.4pt in February, −1.3pt in March, −0.6pt in April, +0.5pt in May, −3.7pt in June and −2.6pt in July — below the prior year in every month except May. Note, however, that from the January 2026 data the survey’s stratification criterion changed from number of employees to number of rooms, and the Japan Tourism Agency itself notes that year-on-year comparisons may include the effect of this revision, so these figures should be discounted somewhat. In 2025 as a whole, monthly values ranged from 51.9% (December) to 66.6% (October), putting Okinawa’s actual occupancy at roughly “around 60% for the year, mid-60s in peak months.”
Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey
*2024 and 2025 are final figures; 2026 are preliminary (up to the second preliminary release). These are actual results based on surveyed properties’ responses, and their definition differs from OTA listed-inventory-based occupancy.
By property type, July 2026 was 79.5% for business hotels, 72.3% for city hotels, 69.5% for resort hotels, 42.0% for ryokan and 32.8% for simple lodgings. Okinawa’s overall figure stays in the 60% range because resort hotels, which hold many rooms, run at around 70%, while ryokan and simple lodgings at 30–40% pull the total down. A year earlier (July 2025) business hotels were at 80.9%, resort hotels at 71.8% and city hotels at 64.6%, so city hotels moved +7.7pt, resort hotels −2.3pt and business hotels −1.4pt.
| Property type | July 2026 (second preliminary) | July 2025 (final) | Difference |
|---|---|---|---|
| Total | 61.8% | 64.4% | −2.6pt |
| Resort hotels | 69.5% | 71.8% | −2.3pt |
| Business hotels | 79.5% | 80.9% | −1.4pt |
| City hotels | 72.3% | 64.6% | +7.7pt |
| Ryokan | 42.0% | 42.4% | −0.4pt |
| Simple lodgings | 32.8% | 40.9% | −8.1pt |
Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey
*Because the stratification criterion changed from number of employees to number of rooms from the January 2026 data, differences from the same month a year earlier may include the effect of the survey design change.
OTA listed-inventory-based occupancy — 86.7% in September, 91.5% for October 1–19, peaking at 96.1% on Saturday, October 10
MetroEngines Research’s OTA listed-inventory-based occupancy, on the other hand, tracks the current month day by day without waiting for the Japan Tourism Agency’s release. The monthly average for all properties in Okinawa was 86.7% in September 2026 (708 properties / 36,220 rooms, 30 days) and 91.5% for the 19 days from October 1 to 19 (933 properties / 37,131 rooms). The aggregation period runs to October 19, and figures for the full month will change with further observation.
By day in October, Saturday, October 10 was the period’s high at 96.1% with an average listed price of ¥38,400, and Thursday, October 1 was the low at 87.2%. October 10 was a weekend where the Naha Great Tug-of-War Festival (October 10–12) coincided with the Sports Day three-day weekend; the following Sunday, October 11, held at 93.2%, but the last day of the holiday, Monday, October 12 (national holiday), fell to 87.5%, and the post-holiday trough continued with 88.0% on Tuesday, October 13. Rooms then filled again toward the weekend, with 93.5% on Thursday the 15th and 94.4% on Friday the 16th; Saturday the 17th was 92.5%, but its average listed price of ¥33,000 was second only to the 10th. Fridays and Saturdays averaged 93.7% (6 days) against 90.5% for other days (13 days), a difference of only 3.2pt.
Source: Compiled by MetroEngines Research and the HotelBank Editorial Team
*OTA listed-inventory-based occupancy (estimate). Covers a fixed cohort of properties listing at least 30% of total rooms on OTAs and similar channels (all properties: 933 properties / 37,131 rooms; business hotels: 140 / 12,924 rooms; resort hotels: 196 / 16,608 rooms). Because unlisted rooms are counted as sold, it runs higher than actual occupancy. Aggregation period: October 1–19, 2026.
| Stay date | All properties | Business hotels | Resort hotels | Average listed price (all properties) | Rooms remaining on OTAs (all properties) |
|---|---|---|---|---|---|
| 10/1 (Thu) | 87.2% | 87.7% | 89.3% | ¥25,424 | 4,763 |
| 10/2 (Fri) | 91.1% | 92.2% | 91.4% | ¥26,926 | 3,324 |
| 10/3 (Sat) | 93.4% | 94.0% | 93.2% | ¥28,572 | 2,392 |
| 10/4 (Sun) | 90.6% | 90.0% | 92.4% | ¥23,705 | 3,375 |
| 10/5 (Mon) | 91.1% | 91.6% | 92.1% | ¥25,276 | 3,214 |
| 10/6 (Tue) | 92.0% | 93.4% | 92.3% | ¥23,263 | 2,864 |
| 10/7 (Wed) | 92.3% | 94.2% | 91.9% | ¥25,011 | 2,764 |
| 10/8 (Thu) | 92.7% | 94.2% | 93.2% | ¥24,330 | 2,593 |
| 10/9 (Fri) | 94.5% | 96.1% | 93.8% | ¥29,734 | 1,982 |
| 10/10 (Sat) | 96.1% | 97.7% | 95.5% | ¥38,362 | 1,420 |
| 10/11 (Sun) | 93.2% | 93.8% | 93.5% | ¥28,255 | 2,456 |
| 10/12 (Mon, holiday) | 87.5% | 84.8% | 90.5% | ¥21,069 | 4,520 |
| 10/13 (Tue) | 88.0% | 85.9% | 91.2% | ¥20,861 | 4,355 |
| 10/14 (Wed) | 90.4% | 90.4% | 92.1% | ¥22,454 | 3,524 |
| 10/15 (Thu) | 93.5% | 95.4% | 93.1% | ¥25,116 | 2,376 |
| 10/16 (Fri) | 94.4% | 96.3% | 93.7% | ¥29,691 | 2,029 |
| 10/17 (Sat) | 92.5% | 94.6% | 91.9% | ¥32,988 | 2,739 |
| 10/18 (Sun) | 88.3% | 88.2% | 89.7% | ¥26,580 | 4,261 |
| 10/19 (Mon) | 90.1% | 92.0% | 90.7% | ¥26,703 | 3,632 |
Source: Compiled by MetroEngines Research and the HotelBank Editorial Team
*Average listed price is the simple average of per-room rates for two guests (tax included). Rooms remaining on OTAs is the listed inventory at the time of observation and excludes unlisted rooms. Aggregation period: as of October 1–19, 2026.
Turning to September, within a monthly average of 86.7%, the first half of the five-day Silver Week holiday (September 19–23) stood out (Saturday, September 19: 92.1%; Sunday the 20th: 95.2%; Monday the 21st, national holiday: 93.1%), while Sunday, September 27 (81.9%) and Monday, September 28 (79.8%), after the holiday, were the month’s low points. Fridays and Saturdays in September averaged 88.7% against 86.0% for other days, so the day-of-week gap was smaller than in October, but the daily range (high 95.2%, low 79.8%, 15.4pt) was larger. In October the floor rose to 87.2%, narrowing the range to 8.9pt, and it is reasonable to read “shallower troughs” as what lifted October’s average.
By segment — business and resort hotels both at 92.2% in October; the difference lies in day-of-week swings
Computing the same indicator by segment, October 2026 (1–19) came out level at 92.2% for business hotels (140 properties / 12,924 rooms) and 92.2% for resort hotels (196 properties / 16,608 rooms). In September business hotels were at 86.7% and resort hotels at 87.7%; in July business hotels were at 90.7% and resort hotels at 90.3%, so the gap in monthly averages has consistently stayed around 1pt. City hotels, with only 8 properties covered, were excluded from this indicator.
Daily movements, however, differ. October business hotels peaked at 97.7% on Saturday the 10th, then fell to 84.8% on Monday the 12th (national holiday) and 85.9% on Tuesday the 13th, a range of 12.9pt. Fridays and Saturdays averaged 95.2% against 90.9% for other days, a day-of-week gap of 4.3pt. Resort hotels ranged from a high of 95.5% (the 10th) to a low of 89.3% (the 1st), a range of 6.2pt, and the gap between Fridays/Saturdays (93.2%) and other days (91.7%) was only 1.5pt. The numbers show the structure: business hotels, centred on Naha city, swing day to day as weekend and holiday tourist demand overlaps with weekday business, family-ceremony and home-visit demand, while resort hotels are smoothed out by multi-night and stay-type bookings. For how business-hotel inventory ran down over this three-day weekend, tracked as booking progress from 45 days before the stay date, see our booking curve analysis of the Naha Tug-of-War three-day weekend.
| Segment (OTA listed-inventory basis) | Coverage | Jul 2026 | Aug 2026 | Sep 2026 | Oct 2026 (1–19) | October range |
|---|---|---|---|---|---|---|
| All properties | 933 properties / 37,131 rooms | 89.3% | 85.5% | 86.7% | 91.5% | 8.9pt |
| Business hotels | 140 properties / 12,924 rooms | 90.7% | 84.7% | 86.7% | 92.2% | 12.9pt |
| Resort hotels | 196 properties / 16,608 rooms | 90.3% | 87.2% | 87.7% | 92.2% | 6.2pt |
Source: Compiled by MetroEngines Research and the HotelBank Editorial Team
*Property and room counts are for the October 2026 cohort. July–September use each month’s cohort (all properties 708–713, business hotels 132–135, resort hotels 174–184). The range is the difference between the daily maximum and minimum.
A 27.5pt gap in the same July — where does the definitional difference come from?
To sum up so far: in July 2026 Okinawa’s official Japan Tourism Agency figure was 61.8% and its OTA listed-inventory-based occupancy 89.3%, a gap of 27.5pt. June was 55.1% vs. 84.9% (a 29.8pt gap) and May 56.7% vs. 83.4% (26.7pt), so a gap of roughly 25–30pt exists consistently regardless of the month. The direction agrees: on both indicators July is higher than May and June.
Most of the gap can be explained by the definitions. First, OTA listed-inventory-based occupancy counts rooms not listed on OTAs as sold. The more rooms a property sells through direct channels, travel agencies, corporate contracts and groups, the smaller its observed remaining inventory and the higher its estimated occupancy. Second, the set of properties differs. Our cohort is limited to properties listing at least 30% of total rooms on OTAs, so most of the ryokan (42.0% in July) and simple lodgings (32.8%) in the Japan Tourism Agency’s population are not included. Third, the Japan Tourism Agency figure is an actual result on an operating-day basis, and closed days and rooms under renovation are also treated differently.
Taking the gap by segment supports this explanation. Business hotels show 79.5% on the Japan Tourism Agency figure against our 90.7%, a gap of 11.2pt; resort hotels show 69.5% against 90.3%, a gap of 20.8pt. Business hotels put a high share of rooms on OTAs, so the two indicators tend to converge. Resort hotels are thought to sell relatively more through travel-agency packages, direct bookings and groups, which produces a picture in which actual occupancy stays around 70% even when remaining OTA inventory is small. OTA listed-inventory-based occupancy therefore cannot be used as a substitute for actual occupancy; the rule is to “compare values measured under the same definition.” For the same July, the national figures were 60.8% from the Japan Tourism Agency (first preliminary; 61.0% in the second preliminary) against 83.5% on the inventory basis, a gap of 22.7pt, so Okinawa’s 27.5pt is larger. The national gap is broken down in our analysis decomposing the 22.7pt gap between the national 60.8% room occupancy and the 83.5% OTA inventory-based figure into four factors.
Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey and MetroEngines Research
*July 2026. Japan Tourism Agency figures are second preliminary; our figures are OTA listed-inventory-based occupancy (estimate; 710 properties / 35,632 rooms; business hotels 132 / 12,490 rooms; resort hotels 174 / 16,351 rooms). Because the definitions differ, the gap mainly reflects “whether sales outside OTAs are included” and “differences in the properties covered,” not “high or low occupancy.”
Relationship with rates — rates are higher in high-occupancy months, but August earns on rate and October on occupancy
Are rates higher in months with higher occupancy? Listing Okinawa’s estimated settled ADR (this article’s ADR, tax-excluded equivalent) by month gives ¥11,200 in May 2026 (N=501), ¥11,400 in June (N=505), ¥13,700 in July (N=508), ¥14,700 in August (N=513), ¥12,000 in September (N=515) and ¥12,800 in October (N=511, an estimate based on listed rates at the time of the survey). OTA listed-inventory-based occupancy was 89.3% in July, 85.5% in August, 86.7% in September and 91.5% in October, so August, with the highest ADR, is below July and October on occupancy. A seasonal structure typical of Okinawa emerges: August, the summer holiday month, captures rate, while October, when holidays coincide with pleasant weather, captures occupancy.
Year on year, estimated settled ADR was up in every month: +7.1% in July, +2.7% in August, +10.0% in September and +18.2% in October (an estimate based on listed rates at the time of the survey). Listed prices (all-plan average, tax included) were ¥31,600 in August (+7.4% YoY), ¥27,800 in September (+11.7%) and ¥27,200 in October (+12.9%). Daily data tell the same story: the correlation coefficient between occupancy and average listed price across all properties was 0.69 for October 1–19 and a high 0.79 for September, confirming that prices rise on days that fill up. The ¥38,400 on Saturday, October 10 is a typical case, where a structure in which higher-priced plans remain listed on high-demand days overlaps with properties’ own pricing.
Source: Compiled by MetroEngines Research and the HotelBank Editorial Team
*Occupancy is on an OTA listed-inventory basis (estimate; each month’s cohort of 589–933 properties). October covers the 1st to 19th. ADR is estimated settled ADR (tax-excluded equivalent, N=501–515 properties); October is an estimate based on listed rates at the time of the survey.
| Month | Japan Tourism Agency room occupancy (Okinawa, total) | OTA listed-inventory-based occupancy (estimate) | Difference | Estimated settled ADR | ADR YoY |
|---|---|---|---|---|---|
| May 2026 | 56.7% | 83.4% | 26.7pt | ¥11,248 | +12.5% |
| Jun 2026 | 55.1% | 84.9% | 29.8pt | ¥11,373 | +15.8% |
| Jul 2026 | 61.8% | 89.3% | 27.5pt | ¥13,722 | +7.1% |
| Aug 2026 | Not yet released (prefecture level) | 85.5% | — | ¥14,652 | +2.7% |
| Sep 2026 | Not yet released | 86.7% | — | ¥12,040 | +10.0% |
| Oct 2026 (1–19) | Not yet released | 91.5% | — | ¥12,828 (estimate) | +18.2% (estimate) |
Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey and MetroEngines Research
*Japan Tourism Agency figures are 2026 preliminary values. Our May figure is for reference only, as observations had accumulated only for the first half of the month. ADR is estimated settled ADR (tax-excluded equivalent); October is an estimate based on listed rates at the time of the survey.
Using them in investment decisions — the Japan Tourism Agency for full-year actual occupancy, the OTA inventory basis for near-term demand strength
The two indicators do not compete; they play different roles. If you need a “full-year occupancy rate” for a business plan, an acquisition underwriting or a loan screening, the figure you can rely on is the Japan Tourism Agency’s official one. Okinawa’s monthly actual occupancy in 2025 ranged from 51.9% to 66.6% overall, 55.0% to 75.0% for resort hotels and 63.7% to 83.4% for business hotels, and the sensible approach is to refer to the property-type figure that matches your own segment. Year-on-year trends and comparisons with other prefectures are also best done with the Japan Tourism Agency figures, which share the same definition. Putting the OTA inventory basis’s “90%” in as a full-year assumption would overstate actual occupancy by 25–30pt.
On the other hand, the Japan Tourism Agency’s prefecture-level figures come out more than two and a half months after the month in question, and first preliminary figures are revised in the second preliminary and final figures. The information needed for operating decisions — whether near-term demand is strong or weak, which segment fills first over a holiday, how large the gap between Fridays/Saturdays and weekdays is — comes faster and in more detail from OTA listed-inventory-based occupancy. For Okinawa in October 2026, the daily shape of 91.5% for the 1st–19th, 96.1% on the first day of the three-day weekend and 87–88% after the holiday is material for thinking about raising and lowering prices and allocating inventory. A realistic sequence for prospective entrants is to grasp “the level of full-year actual occupancy” from the Japan Tourism Agency figures, then check “the current momentum of demand and the differences between segments” on the OTA inventory basis.
Some supply-side context is also worth adding. In Naha city, room supply from converting existing buildings continues, and, as covered in our analysis of the two-building, 342-room conversion at Naha Asahibashi, supply that does not depend on new construction affects business-hotel supply and demand in Naha. Okinawa also draws school-trip demand in December–January as an off-peak destination, so even in winter, when occupancy looks low on the Japan Tourism Agency figures, some segments have a demand peak. Which definition of occupancy you look at is a premise for these discussions of supply and seasonality.
⚠ Note on ADR for future dates: ADR (estimated settled ADR) and listed prices for future dates in this article are based on selling prices published on OTAs at the time of the survey and will change as the check-in date approaches. Please note that prices currently set high may fall with last-minute discounts. OTA listed-inventory-based occupancy for October 2026 is likewise an interim reading as of October 1–19, and figures for the full month will change with further observation.
Frequently asked questions
Q. Where can I find Okinawa’s occupancy rate?
A. The official actual figures are the room occupancy rates by prefecture and property type in the Japan Tourism Agency’s Overnight Travel Statistics Survey; the tabulated results (Excel) and time-series tables are available on the Japan Tourism Agency’s website. The latest figure for Okinawa is 61.8% for July 2026 (second preliminary). Prefecture-level figures are released one month after the national first preliminary figures. For near-term demand strength, there are methods that estimate from listed OTA inventory, such as the OTA listed-inventory-based occupancy (estimate) shown in this article.
Q. How much are the first preliminary figures revised?
A. The Japan Tourism Agency’s first preliminary figures are based on survey forms collected so far and are revised in the second preliminary figures and then in the annual final figures. In our verification, room occupancy moves by about 0.6pt on average from the first to the second preliminary figures (for details, see What Is Japan’s Accommodation Survey? First Estimates Move 0.6pt). In addition, from the January 2026 data the stratification criterion changed from number of employees to number of rooms, so year-on-year comparisons with 2025 may include the effect of the survey design change.
Q. How large is the gap between OTA inventory-based occupancy and actual occupancy?
A. In Okinawa it was 26.7–29.8pt in May–July 2026, and 27.5pt in July (61.8% from the Japan Tourism Agency vs. our estimate of 89.3%). By segment, the gap was 11.2pt for business hotels and 20.8pt for resort hotels: the larger the share of sales channels outside OTAs a segment is thought to have, the wider the gap. The OTA inventory-based figure is an estimate that counts unlisted rooms as sold; rather than substituting it for actual occupancy, it is suited to comparing values under the same definition across adjacent months, days of the week and segments.
Summary
Okinawa’s occupancy differs by 25–30pt between the Japan Tourism Agency’s official figure (61.8% in July 2026, second preliminary) and the OTA listed-inventory-based estimate (89.3% in the same month; 91.5% for October 1–19), and the gap arises mainly from the definitions: “whether sales outside OTAs are included” and “differences in the properties covered.” By segment, the gap differs — 11.2pt for business hotels and 20.8pt for resort hotels — and actual resort occupancy stays around 70%. For operators investing in or entering Okinawa, the realistic reading is to use the Japan Tourism Agency figures for full-year actual occupancy and business-plan assumptions, and the OTA inventory basis for near-term demand strength and day-of-week and segment differences. In October, Okinawa rose to 96.1% with an average listed price of ¥38,400 on the first day of the three-day weekend, and did not fall below 87% even after the holiday. Estimated settled ADR was also above the prior year in every month from July to October 2026 (+2.7% to +18.2%), so demand strength is continuing on both occupancy and rate.
Related articles
- What Is Japan’s Accommodation Survey? First Estimates Move 0.6pt
- Overnight Travel Statistics Survey Guide: 61.6% OCC, Prelim vs Final
- Naha Asahibashi: 2 Building Conversions Add 342 Rooms; ADR ¥10,300
- Off-Peak School Trips: Kyoto, Nara, Nagasaki Jun-Jul, Okinawa Dec-Jan
References and Sources
■ Data sources
The Japan Tourism Agency’s Overnight Travel Statistics Survey time-series tables (Okinawa; 2024 and 2025 final, January–July 2026 preliminary) and the July 2026 second preliminary tabulated results (by property type). MetroEngines Research’s OTA listed-inventory-based occupancy (Okinawa, May 2026 to October 19, 2026; each month’s fixed cohort of 589–933 properties), estimated settled ADR and listed prices (N=501–515 properties).
■ Calculation assumptions
OTA listed-inventory-based occupancy = 100 − 100 × rooms remaining on OTAs ÷ total rooms. Limited to properties listing at least 30% of total rooms on OTAs and similar channels, using observations within 45 days before the stay date. Monthly averages are simple averages of daily values. The gap (pt) against the Japan Tourism Agency figure is the simple difference for the same month; by segment, business hotels were compared with business hotels and resort hotels with resort hotels.
■ Limitations and caveats
The Japan Tourism Agency’s 2026 values are preliminary and may be revised, and because the stratification criterion changed from the January 2026 data, year-on-year comparisons may include the effect of the survey design change. OTA listed-inventory-based occupancy counts unlisted rooms as sold, so it runs higher than actual occupancy and cannot be used in its place. October is an interim reading for the 1st–19th, and October’s estimated settled ADR is an estimate based on listed rates at the time of the survey. City hotels were excluded from the segment breakdown because only 8 properties were covered.
- Japan Tourism Agency, “Overnight Travel Statistics Survey (August 2026 first preliminary, July 2026 second preliminary)” (September 30, 2026; in Japanese)
- Japan Tourism Agency, “Overnight Travel Statistics Survey” tabulated results and time-series tables (July 2026 second preliminary results Table 8; time-series tables 4-1 and former 4-2; in Japanese)
- Japan Tourism Agency, “Overnight Travel Statistics Survey: Change to the Stratification Criteria” (in Japanese)
- City of Naha, “The 56th Naha Great Tug-of-War Festival” (October 10–12, 2026; in Japanese)
- MetroEngines Research — Okinawa OTA listed-inventory-based occupancy (fixed cohort; October 2026: 933 properties / 37,131 rooms), estimated settled ADR (N=501–515 properties), listed prices (two guests per room, tax included)
MetroEngines Research tracks about 168,000 properties in Japan, of which about 27,000 properties (about 1.26 million rooms) confirmed as operating on OTAs form the analysis universe. The property and room counts in this article are Okinawa figures within that universe.
Recommended reading
- Japan Hotel Occupancy July 2026: 60.8% Official vs 83.5% OTA Estimate
- Naha Tug-of-War Weekend 2026: Okinawa Business Hotels 97.7% to 84.8%
- Japan Lodging Stats: Preliminary Figures Move 3.4%, Occupancy 1.0pt
- Double Ten x Hangul Day Oct 9-11: Okinawa, Fukuoka Friday Pace +2pt
- Double Ten 2026 x Japan’s Long Weekend: Okinawa 90.8% OCC on Oct 10