How many rooms does a hotel have on average? Public statistics give a clear answer. According to the Ministry of Health, Labour and Welfare (MHLW) “Report on Public Health Administration and Services, FY2025” (衛生行政報告例; as of end-March 2026, released October 27, 2026), Japan has 56,773 ryokan and hotel business facilities with 1,797,781 rooms under the Inns and Hotels Act. That works out to an average of 31.7 rooms per facility, and 14.5 rooms per 1,000 people when divided by the total population of 123.80 million in the Ministry of Internal Affairs and Communications (MIC) “Population Estimates” (as of October 1, 2024). By prefecture, this “depth of capacity for lodging demand” ranges more than tenfold, from 47.2 rooms in Okinawa to 3.7 rooms in Saitama. This article maps and tabulates room density across all 47 prefectures, then overlays MetroEngines Research’s September 2026 estimated occupancy and estimated settled ADR together with guest-night figures from the Japan Tourism Agency’s “Overnight Travel Statistics Survey” to identify which prefectures still have room for new supply, and where lodging demand from prefectures surrounding major cities is going.

The National Picture — 31.7 Rooms per Facility, 14.5 Rooms per 1,000 People

Let us first set out the national outline. At the end of FY2025, there were 56,773 ryokan and hotel business facilities with 1,797,781 rooms. Compared with the previous year-end (52,946 facilities, 1,780,679 rooms), facilities rose +7.2% and rooms +1.0%, with facility growth far outpacing room growth. As a result, the average number of rooms per facility fell by about two, from 33.6 to 31.7. Business permits issued during the year totaled 10,012 nationwide (7,720 the year before) against 3,489 closures, which reads as a year in which small-scale permits broadened the base of supply. The net-supply structure for the previous year (FY2024), breaking down 7,720 permits and 2,817 closures by prefecture and counting the prefectures where rooms declined on a net basis, is analyzed in detail in 98,338 Lodging Facilities: 30 of 47 Prefectures Lost Rooms in FY2024.

When asked for “the average number of rooms in a hotel,” this figure of 31.7 rooms is the benchmark in public statistics. Note, however, that it is an average of facilities counted by permit unit under the Inns and Hotels Act, and includes small ryokan, whole-house rentals and facilities that switched from simple lodging to ryokan/hotel business. This does not contradict the fact that chain limited-service hotels have 100–200 rooms; the breadth of the population simply pulls the average down. The average size by prefecture is examined later, but it differs more than fourfold, from 73.7 rooms in Osaka to 16.1 rooms in Okinawa, so the “average” looks completely different depending on the region.

Rooms per 1,000 people (14.5), on the other hand, is a yardstick for comparing the depth of capacity for lodging demand across prefectures. Populous prefectures also generate large in-prefecture business-trip and homecoming demand, so normalizing by population rather than using absolute room counts reveals how far tourism-oriented prefectures are structured to absorb “demand from outside.”

Room Density in 47 Prefectures — High in Okinawa 47.2, Yamanashi 34.4, Nagano 31.9; Low in Saitama 3.7, Kanagawa 6.4, Nara 8.0

The top prefectures for rooms per 1,000 people are Okinawa at 47.2 (69,200 rooms), Yamanashi at 34.4 (27,239 rooms) and Nagano at 31.9 (63,474 rooms), followed by Ishikawa at 26.3 and Hokkaido at 23.8. All are regions driven mainly by tourism demand from outside the prefecture and overseas, and Okinawa’s density is more than three times the national average. Ishikawa ranks fourth, ahead of Hokkaido, likely because the hot spring areas of Noto and Kaga and the business and tourism demand of Kanazawa are large relative to its population.

At the bottom are Saitama at 3.7, Kanagawa at 6.4, Nara at 8.0, Hyogo at 8.8 and Aichi at 9.1, with Chiba (10.0), Ibaraki (10.2) and Shiga (10.9) also below the national average. Tokyo has the largest room count in Japan at 219,461 rooms, but divided by a population of over 14 million it comes to 15.5 rooms, close to the national median (15.9). Osaka is also near the national average at 14.6, showing that big cities do not have “many rooms” so much as “many people.”

Source: Compiled by HotelBank Editorial Team from MHLW “Report on Public Health Administration and Services, FY2025” (as of end-March 2026) and MIC “Population Estimates” (as of October 1, 2024)

On the map, high-density prefectures are found in Okinawa, Koshin (Yamanashi, Nagano), Hokuriku (Ishikawa, Fukui), Hokkaido, the Sea of Japan side of Tohoku and Oita in Kyushu — all tourism and hot spring destinations. Low-density prefectures are concentrated in the commuter belts of the Tokyo and Kinki metropolitan areas and in Aichi in the Chukyo area; the more a prefecture serves as a bedroom community for a big city, the fewer “places to stay” it has. This is not because demand is small, but because rooms in the neighboring big city absorb that demand — a structure confirmed later with guest-night data.

Circle size = rooms per 1,000 people (placed at each prefectural capital). Source: Compiled by HotelBank Editorial Team from MHLW “Report on Public Health Administration and Services, FY2025” (as of end-March 2026) and MIC “Population Estimates” (as of October 1, 2024).

Room Density × September Estimated Occupancy — Supply Headroom Where Density Is Low but Occupancy High

Room depth on its own says nothing about supply and demand. We therefore built a scatter plot with room density on the horizontal axis and MetroEngines Research’s September 2026 estimated occupancy (based on OTA-listed inventory, estimated) on the vertical axis. It covers 46 prefectures (11,747 properties / 755,381 rooms listed on major booking sites); facility and room counts by prefecture are shown in the table at the end of this article. The correlation coefficient between the two is 0.05, showing almost no tendency for prefectures with more rooms to have lower — or higher — occupancy. September occupancy appears to be determined less by density than by each prefecture’s seasonality of demand and visitor mix.

Of particular note are prefectures whose density is below the national median (15.9 rooms) but whose occupancy is above the median (84.6%). These are Aichi (density 9.1, occupancy 90.4%), Tokyo (15.5, 89.8%), Hiroshima (12.3, 88.5%), Fukuoka (13.1, 88.3%), Gifu (14.3, 88.1%), Miyagi (14.6, 87.9%), Kochi (15.7, 86.9%), Hyogo (8.8, 86.3%), Shiga (10.9, 85.7%), Okayama (11.8, 85.1%) and Saitama (3.7, 84.7%) — all in a state where “there are few places to stay relative to population, yet the rooms that exist filled well in September.” Aichi is in the bottom five for density at 9.1 but among the top nationally for occupancy at 90.4%, likely reflecting overlapping business demand in central Nagoya and demand related to the 2026 Asian Games. Fukuoka (13.1, 88.3%) and Hiroshima (12.3, 88.5%) sit in the same quadrant, highlighting the high occupancy of regional hub cities.

Conversely, even high-density prefectures posted high September occupancy: Okinawa 86.7%, Nagano 88.4% and Hokkaido 90.3%. In these prefectures, even though rooms are plentiful relative to population, the absolute number of visitors is larger still. Yamanashi ranks second in density but its occupancy of 82.2% is slightly below the median, as the seasonality of Fuji Five Lakes and Isawa demand — concentrated in autumn holidays and the foliage season — shows in the September figure. Aomori, at 16.6 rooms, sits near the median in density but posted the highest occupancy of the 46 prefectures at 95.5%, absorbing September tourism and business demand with a limited 19,283 rooms.

Source: Compiled by MetroEngines Research and HotelBank Editorial Team (September estimated occupancy covers 46 prefectures, 11,747 properties / 755,381 rooms, based on major booking-site listings) / Room density from MHLW “Report on Public Health Administration and Services, FY2025” and MIC “Population Estimates”

The upper-left of the scatter plot — the “low density, high occupancy” quadrant — is an environment where existing properties can more easily command rates, and a market where new supply has room to deepen capacity. That said, estimated occupancy is a single-month September figure, estimated from the sell-through of inventory sold on OTAs, and differs from each property’s actual overall occupancy. Judgments on supply headroom should combine year-round occupancy with the scale of guest nights discussed below.

Room Density × Estimated Settled ADR — Little Evidence That Thinner Prefectures Command Higher Rates

Next, keeping room density on the horizontal axis, we switch the vertical axis to September 2026 estimated settled ADR (finalized month, tax-excluded equivalent, area median). It covers 47 prefectures and 17,136 properties in total, and the correlation coefficient of 0.17 is weak here too. The national median is ¥9,700; the highest are Kyoto at ¥14,200 (N=604), Kanagawa at ¥14,200 (N=579), Oita at ¥14,000 (N=470) and Tokyo at ¥13,700 (N=1,139). Kanagawa ranks 46th in density yet second nationally in ADR, with rates supported by a structure in which resort and ryokan demand in Hakone, Yugawara and Kamakura, plus city demand in Yokohama, converge on rooms that are few relative to population.

Among high-density prefectures, Okinawa (¥12,000), Nagano (¥11,800), Yamanashi (¥11,700) and Ishikawa (¥12,400) are ¥2,000–2,700 above the national median, so rates hold up even where rooms are plentiful. Hokkaido (¥11,500) is also above the median. Osaka, on the other hand, fell sharply to ¥9,300 from ¥13,800 in the same month a year earlier. The natural reading is that September rates boosted by Expo 2025 Osaka, Kansai are returning to normal levels the following year — not a change in density itself (14.6 rooms; −0.2% year on year).

Source: Compiled by MetroEngines Research and HotelBank Editorial Team (estimated settled ADR for September 2026, finalized month, 47 prefectures, N=17,136 properties) / Room density from MHLW “Report on Public Health Administration and Services, FY2025” and MIC “Population Estimates”

What this shows is that prefecture-level room density is a weak explanatory variable for rates; rates are determined by “the quality of demand coming into the prefecture.” Factors such as the share of leisure travel, the share of foreign guests and the presence of destination resorts weigh more heavily than rooms per capita. In practice, room density is best used as a measure of supply depth, while rates and occupancy should be read together with demand-side indicators.

The Structure of Commuter-Belt Prefectures — Lodging Demand in Saitama, Kanagawa, Chiba and Nara Flows to Neighboring Big Cities

Why do commuter-belt prefectures of the Tokyo and Kinki areas sit at the bottom for room density? We check using guest nights from the Japan Tourism Agency’s “Overnight Travel Statistics Survey” 2025 annual figures (final). Annual guest nights per resident were 7.61 in Tokyo, 13.16 in Kyoto and 6.71 in Osaka, versus 0.74 in Saitama, 2.90 in Kanagawa, 2.47 in Nara and 4.52 in Chiba — the commuter-belt prefectures are only a fraction of the big cities. Saitama’s 5,403,310 guest nights are one-twentieth the scale of Tokyo’s for a prefecture of about 7.33 million people, consistent with its thin 3.7 rooms per 1,000 people.

This structure does not mean that “there is no demand in the commuter belt.” Many business travelers and visitors to Saitama and Kanagawa stay in rooms in central Tokyo, 30 minutes to an hour away by rail, and travel during the day. Nara is similar: of its 3,177,210 guest nights, 77.5% were at facilities where leisure guests make up 50% or more — the highest in Japan, making it a tourism-oriented prefecture — yet guest nights per resident are only 2.47. Compared with 13.16 in Kyoto and 6.71 in Osaka, the numbers show that tourists visiting Nara are placing their stays on the Kyoto and Osaka side.

Chiba is an exception among the commuter-belt prefectures, with a high 4.52 guest nights per resident, and its 450 annual guest nights per room rival Tokyo’s (492). This is because two anchors for stays — Tokyo Disney Resort and Narita Airport — have the power to keep overnight stays within the prefecture, a character reflected in the 61.5% share of guest nights at facilities where leisure guests make up 50% or more. Its room density of 10.0 is below the national average, but turnover per room is on par with big cities, making it an example of supply keeping up with demand even among commuter-belt prefectures.

Source: Compiled by HotelBank Editorial Team from Japan Tourism Agency “Overnight Travel Statistics Survey,” January–December 2025 (final annual figures, all facilities), MHLW “Report on Public Health Administration and Services, FY2025” and MIC “Population Estimates”

Source: Compiled by HotelBank Editorial Team from Japan Tourism Agency “Overnight Travel Statistics Survey” 2025 annual figures (final), MHLW “Report on Public Health Administration and Services, FY2025” and MIC “Population Estimates”

The flow is just as clear in the foreign guest share: 55.7% in Tokyo, 55.2% in Kyoto and 41.7% in Osaka, versus 5.2% in Saitama, 16.4% in Nara, 18.4% in Kanagawa and 15.5% in Chiba — inbound stays are concentrated almost entirely on the big-city side. For commuter-belt prefectures, the more central-city rooms run at high occupancy and high rates, the more room there is to offer the option of “staying just short of the city center.” Kanagawa’s second-place national ranking in estimated settled ADR is evidence that its rooms, thin relative to population, are supported by strong demand, positioning it as an area with the potential to deepen capacity.

Average Facility Size — Osaka 73.7, Kyoto 52.1, Miyagi 51.0, Okinawa 16.1 Rooms

Looking by prefecture at the frequently searched “average number of rooms in a hotel,” Osaka stands out at 73.7 rooms per facility, followed by Kyoto at 52.1, Miyagi at 51.0, Chiba at 50.7 and Aichi at 47.4. Osaka has 128,218 rooms across 1,739 facilities, dominated by large limited-service and full-service city hotels. Kyoto has a modest 965 facilities, but because 3,890 simple-lodging facilities are counted separately, the facilities that remain in the ryokan/hotel category show a large average size. How prefectural supply rankings change when simple lodging is included in the denominator is explored in Lodging Supply Denominator: 44,901 Simple Lodgings Move Kyoto 21st→5th.

The smallest average sizes are in Okinawa (16.1 rooms), Fukui (18.9), Yamanashi (20.5) and Nagano (21.6), all of which overlap with the top prefectures for room density. In tourist areas with many pensions, small ryokan and condominium-style properties, rooms are spread across “many small facilities,” so facilities are small even where density is high. In Okinawa, facilities grew sharply by +13.3% year on year (3,799 → 4,304), outpacing room growth of +2.1%, so the average size shrank further from 17.8 to 16.1 rooms.

Tokyo is moving in the same direction. Facilities rose +39.9% from 5,858 to 8,196, while rooms rose only +3.1%, shrinking the average size by about 10 rooms, from 36.3 to 26.8. Business permits issued during the year totaled 2,602, a quarter of the national total, indicating that small facilities continued to obtain permits. Fukuoka (facilities +26.8%) and Aichi (+17.9%) showed similar movements; in big cities, the phase is not “large hotels increasing and raising the average” but “small facilities increasing and lowering the average.” This is also a sign that the base of the lodging market is widening, and can be seen as a year in which the door to entry opened wider for small operators.

Source: Compiled by HotelBank Editorial Team from MHLW “Report on Public Health Administration and Services, FY2025” (as of end-March 2026). Top 10 and bottom 8 prefectures shown; dotted line = national average

Table of All 47 Prefectures — Ranked by Room Density

The table below summarizes room density, rooms, facilities, average rooms per facility, September estimated occupancy and September estimated settled ADR for all 47 prefectures. Estimated occupancy is based on major booking-site listings; estimated settled ADR is the area median for the finalized month (September 2026), with the number of covered properties in parentheses. Kagoshima’s September estimated occupancy was excluded from this article’s tabulation.

Source: Compiled by HotelBank Editorial Team from MHLW “Report on Public Health Administration and Services, FY2025” (as of end-March 2026), MIC “Population Estimates” (as of October 1, 2024) and MetroEngines Research (September 2026, estimates)

Implications for Operators and Investors — Use Room Density as a Measure of Capacity Depth

First, prefectures with low room density and high occupancy are markets where existing properties have ample room to build rates. In prefectures such as Aichi, Hiroshima, Fukuoka, Hyogo and Okayama, where rooms are few relative to population and September occupancy exceeds the median, and in prefectures such as Kanagawa, where thin room stock fills at the second-highest rates in Japan, there is a revenue opportunity to raise peak-season prices in stages. An effective approach is to watch the booking curve and raise prices in line with the remaining-inventory trend within a 45-day lead time.

Second, for those considering new supply, thin room density can be a sign that “capacity is insufficient,” but in commuter-belt prefectures the structure of demand flowing out to the neighboring big city must be factored in. As Chiba shows, if a prefecture has anchors that keep stays within it (theme parks, airports, hot spring areas), big-city-level turnover of 450 guest nights per room per year is achievable. For sites without such anchors, a realistic strategy is pricing designed to capture “stay-short-of-the-city” demand during high-rate periods in the city center. For the prefecture-level supply impact of rooms opening in 2026 relative to existing stock, Japan Hotel Supply 2026: Saitama 8.0%, Kumamoto 7.9% of Existing Roomsis also a useful reference.

Third, “the average number of rooms in a hotel” ranges from 16 to 74 depending on the region, so competitive comparisons and development-scale planning should use each prefecture’s average size rather than the national average of 31.7 rooms. In big cities where permits for small facilities have increased, properties of a few dozen rooms are multiplying in the same trade areas, pointing existing large hotels toward differentiation in segments where economies of scale work, such as groups, MICE and long stays.

Summary

Measured by public statistics, Japan’s lodging supply at the end of March 2026 consisted of 56,773 ryokan and hotel business facilities with 1,797,781 rooms — an average of 31.7 rooms per facility and 14.5 rooms per 1,000 people. Room density is high in tourism-oriented prefectures such as Okinawa, Yamanashi and Nagano, and low in big-city commuter-belt prefectures such as Saitama, Kanagawa and Nara. Correlations between density and September estimated occupancy and estimated settled ADR are both weak; prefecture-level room depth does not directly determine supply-demand balance or rates. Thin capacity in commuter-belt prefectures is the result not of absent demand but of outflow to neighboring big cities, and areas such as Kanagawa, where thin room stock fills at high rates, retain the potential to deepen capacity. Average size is large in Osaka, Kyoto and Miyagi and small in Okinawa, Koshin and Hokuriku. In big cities, the average size is shrinking as small facilities increase, indicating a phase in which the base of the lodging market is widening.

Frequently Asked Questions

Q. What is the average number of rooms in a hotel?
According to the MHLW “Report on Public Health Administration and Services, FY2025” (as of end-March 2026), there are 56,773 ryokan and hotel business facilities with 1,797,781 rooms under the Inns and Hotels Act, for an average of 31.7 rooms per facility. By prefecture, Osaka is the largest at 73.7 rooms and Okinawa the smallest at 16.1 rooms.

Q. Which prefectures have the most rooms per capita?
In rooms per 1,000 people, Okinawa (47.2), Yamanashi (34.4) and Nagano (31.9) lead. The national average is 14.5; Tokyo is at 15.5 and Osaka at 14.6, close to the national average.

Q. Why do Saitama and Nara have so few rooms?
Because lodging demand flows to neighboring Tokyo, Kyoto and Osaka. In the Japan Tourism Agency’s final 2025 figures, guest nights per resident were 0.74 in Saitama vs. 7.61 in Tokyo, and 2.47 in Nara vs. 13.16 in Kyoto, showing that visitors place their stays on the big-city side.