Capsule hotels have become the lodging type whose inventory fills most easily. In MetroEngines Research data for August 2026, the occupancy rate (OTA-listed-inventory basis, estimated; hereafter the “inventory sell-through rate”) of 27 capsule hotels in Tokyo was 96.3%, the highest of eight property types, ahead of business hotels (91.1%, N=872), city hotels (90.7%, N=100) and deluxe hotels (93.1%, N=40). In all 11 prefectures with three or more capsule hotels, capsules beat the all-property average. However, this figure measures how fully the inventory placed on OTAs sold, and many capsule hotels list only a small fraction of their total beds on OTAs. This article first plots sell-through and price for the eight property types on the same axes, then corrects this surface reading using the OTA listing ratio, and finally uses the geographic distribution of 237 capsule hotels (about 30,000 beds) nationwide and the opening pipeline to identify which cities and price bands still have room to grow.

Sell-Through × Price for 8 Property Types — Capsules Alone Sit in the “Cheapest, Fullest” Quadrant

Taking Tokyo as the example, we first plot the eight property types on two axes: sell-through rate and listed price. In August 2026, Tokyo’s 1,383 properties (184,093 rooms) had a sell-through rate of 91.0% and a listed price (all-plan average, 2 guests per room, tax included) of ¥20,900. By comparison, business hotels posted 91.1% at ¥16,600, city hotels 90.7% at ¥33,500, deluxe hotels 93.1% at ¥59,800 and ryokan 91.9% at ¥26,900. Even with a more than threefold price spread, sell-through stayed within a 90–93% band. The simple relationship that higher prices fill less easily was not observed in Tokyo in August.

Capsule hotels break above that band. Across 27 properties and 4,115 beds, sell-through was 96.3% and the listed price was ¥11,900, the lowest of the eight types. Capsules are the only type in the upper-left “cheap and full” quadrant of the scatter chart, 5.2 points above business hotels. The type that breaks below is resort (82.7%, N=14), but only 14 such properties are in Tokyo, so composition effects are large.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Tokyo, August 2026; bubble size = number of properties)

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Tokyo, August 2026. Luxury is a cross-cutting segment of properties with a listed price of ¥100,000 or more and overlaps with other types)

Tracking daily data makes the nature of the gap clearer. In Tokyo in August, the all-property sell-through rate bottomed at 85.5% on Sunday, August 16 and Monday, August 31, after the Obon holidays, and business hotels fell to 84.6% on the same day. Capsule hotels’ monthly low was 94.0% on Monday, August 17, and their high was 98.5% on Saturday, August 29. In a month when all properties swung by nearly 10 points, capsules moved within a range of just 4.5 points. Staying full even in demand troughs is what created the gap in the monthly average.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Tokyo, August 2026, daily. Capsule N=27, business N=872, all properties N=1,383)

Above the All-Property Average in All 11 Prefectures — But Only 3–5 Properties Each Outside Tokyo

To check whether this is a Tokyo-only phenomenon, we extended the same calculation to all 47 prefectures. Nationwide, 80 capsule hotels have a computable sell-through rate, and 11 prefectures contain three or more. In all 11, capsules beat the all-property average without exception: the gap was smallest in Kumamoto at +4.6 points, +5.3 points in Tokyo, and largest in Osaka at +12.1 points (98.0% vs. 85.9%). Weighted by beds, the sell-through rate of the 62 capsule hotels (8,552 beds) in these 11 prefectures is 97.1%.

Outside Tokyo’s 27 properties, however, samples are small at 3–5 properties. Fukuoka (5 properties) posted 98.9%, Osaka (5) 98.0%, Hokkaido (4) 98.1%, Kanagawa, Chiba and Aichi (3 each) in the 98% range, and Kyoto (3) 94.8%; all should be read as indicative given the small samples. A change in one property’s inventory settings could move these by several points, so the table shows N for each prefecture. Even so, the fact that the direction is consistent across 11 regions supports the view that Tokyo’s result is not a coincidence.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (August 2026; 11 prefectures with 3+ capsule hotels. Outside Tokyo, small samples; indicative only)

Source: MetroEngines Research; compiled by HotelBank Editorial Team (August 2026. Prices are listed prices, all-plan average, 2 guests per room, tax included. Capsules outside Tokyo are small samples; indicative only)

Shifting the month does not change the ranking. Tokyo’s 27 capsule hotels posted 97.5% in July 2026 (all properties: 93.2%) and 97.4% in September (all properties: 92.2%), ranking first of eight types for three consecutive months. Capsules are not a type that fills only in the summer peak; the fact that sell-through did not fall even in September, as the market heads into the off-season, feeds into the discussion of supply headroom below.

Correcting the Surface Reading with the OTA Listing Ratio — Capsules List Only 13% of Their Beds on OTAs

Here we need to pause. The sell-through rate is an estimate that restates “the share of inventory listed on OTAs that sold” against total rooms, and its definition is entirely different from the room occupancy rate in the Japan Tourism Agency’s “Overnight Travel Statistics Survey” (64.4% overall in the first preliminary figures for August 2026). The Japan Tourism Agency statistics count actual occupancy of all rooms reported by properties, while this article’s metric is back-calculated from remaining OTA availability. Rooms not listed on OTAs are counted as “not remaining,” so the lower a property’s OTA listing ratio, the higher its sell-through rate reads. We must test whether part of the reason capsules appear at the top is this surface effect. Where estimated OCC and sell-out rate diverge is examined through booking curves for three Tokyo property types in our article Estimated OCC vs. Sell-Out Rate: Tokyo Business 77.7% vs. Capsule 93.8%.

We therefore calculated the “OTA listing ratio” — the maximum inventory each property listed on OTAs during the observation period divided by its total beds — for 62 capsule hotels in Tokyo, Osaka, Fukuoka and Hokkaido and, for comparison, 100 Tokyo business hotels (the 100 with the lowest remaining-room ratio among 483 with data for August 20, 2026 check-ins). The result was clear: capsules had a median listing ratio of 13.0% (14.3% for the 39 in Tokyo alone), versus 45.1% for business hotels. Of the 62 capsules, 27 (43.5%) listed less than 10%, compared with only 12 of 100 business hotels. On the other hand, 19 capsule hotels (30.6%) listed 50% or more, so the distribution is polarized between a “list very thinly” group and a “list most beds” group. The structure in which, regardless of type, rooms placed on booking sites make up only about half of total rooms is laid out across 11,000 properties in OTAs Get 51.8% of Rooms at 11,152 Hotels — the ¥19M Staffing Upside.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Capsules: 62 properties in Tokyo, Osaka, Fukuoka and Hokkaido; business: 100 properties in Tokyo. Observed for August 20, 2026 check-ins)

This polarization has reasons specific to the format. In a capsule hotel, one room equals one bed, with no hierarchy of room types such as twins or suites. Because inventory cannot be allocated by price tier, the main inventory control lever is allotment: how many beds to put on OTAs. When a 160-bed property lists only 10–20 beds on OTAs and replenishes them as they sell, remaining OTA availability often touches zero. This practice pushes the sell-through rate up. However, public data alone cannot tell whether direct sales via the official website or corporate contracts dominate, whether OTA allotments are contractually kept small, or whether inventory is released in small increments according to sales pace; several of these likely coexist.

There is one way to check the size of the surface effect in numbers. Rebuilding August in Tokyo using only properties with an OTA listing ratio of 30% or more lowers the all-property sell-through rate by 2.8 points from 91.0% to 88.2%, and business hotels go from 91.1% to 88.8%. Under this condition, not a single capsule hotel remains. In other words, none of the 27 consistently listed 30% or more of total beds throughout the month, so the most accurate reading of 96.3% is not actual capsule occupancy but an upper bound showing that “thinly listed inventory repeatedly sold out.”

Does the conclusion that capsule supply and demand is tight then disappear? We think not, for three reasons. First, the low listing ratio is common to July and September 2026 as well, yet the gap versus all properties held in September (+5.2 points) rather than only in August with its Obon holidays. Second, even including the 19 “list most beds” properties with listing ratios of 50% or more, capsules’ daily sell-through low was 94.0%, 8.5 points above the all-property low of 85.5%. Third is price. Tokyo capsules’ estimated settled ADR was ¥5,400 in August 2026 (N=39, per bed, tax-excluded equivalent), up +4.6% from ¥5,200 (N=41) a year earlier. Given that Tokyo’s all-property ADR in the same month was ¥11,800 (N=1,140), down −9.2% from ¥13,000 a year earlier, capsules were filling while raising rates in a month when the market gave up rate. A price-side signal that does not depend on how inventory is listed backs up the strength of supply and demand.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (Tokyo capsules, estimated settled ADR, per bed, tax-excluded equivalent. 2024: N=57–59; 2025: N=29–60; 2026: N=32–39. January 2026 and November–December 2025, when the number of properties fell below 60% of the series median, are not shown)

Osaka capsules’ ADR dropped sharply to ¥4,000 (N=10) in August 2026 from ¥5,600 a year earlier, and Fukuoka was ¥5,700 (N=6), down −6.5% from ¥6,100. Osaka’s prior-year month fell during the Expo, an unusual comparison, and both have N of 10 or less, so we note them here as indicative only.

Geographic Distribution of 237 Properties and 30,000 Beds — 41% in Tokyo, 69% in Four Prefectures

Where is supply for this tight segment located? Within MetroEngines Research’s coverage, 237 capsule hotels with 30,453 beds were confirmed operating nationwide as of October 2026. By prefecture, Tokyo accounts for 94 properties and 12,516 beds, or 41.1% of the total, followed by Osaka (26 properties, 4,697 beds), Kyoto (16 properties, 1,856 beds) and Kanagawa (15 properties, 1,875 beds). These four prefectures hold 68.8% of beds, and capsule hotels are located in only 31 prefectures. In 16 prefectures, not a single capsule hotel could be confirmed operating on OTAs.

At the municipal level the concentration is even stronger. The top areas are Osaka’s Chuo Ward (14 properties, 2,687 beds), Tokyo’s Minato Ward (14 properties, 1,755 beds), Taito Ward (13 properties, 2,097 beds), Chuo Ward (11 properties, 1,595 beds) and Shinjuku Ward (10 properties, 1,736 beds), clustered within walking distance of entertainment districts and rail terminals. Yokohama’s Naka Ward (8), Fukuoka’s Hakata Ward (8), Sapporo’s Chuo Ward (6), and Kyoto’s Nakagyo Ward (6) and Shimogyo Ward (5) follow the same pattern.

By size, properties with 100–199 beds are the most common at 101 (42.6%), while the 37 with 200 or more beds total 10,600 beds, or 34.8% of all beds. There are also 37 small properties with fewer than 50 beds, but they account for only 970 beds, or 3.2%. Mid- to large-size properties form the core of the stock, and the average of 128 beds per property is close to the average room count of a business hotel.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (237 capsule hotels confirmed operating, as of October 2026)

Opening Pipeline and Construction Starts — After the 2017–19 Wave, a Quiet Increase of About 10 a Year

Tracking when supply accumulated by opening year shows a clear wave in capsule supply. Openings (confirmed via OTA listings) went from 12 in 2015 to 24 in 2016, 45 in 2017, 42 in 2018 and 36 in 2019; over these five years, which coincided with the first inbound wave, 159 properties with 17,762 beds opened. More than half of today’s stock was born in this period. Openings fell to 16 in 2020 and 7 each in 2021 and 2022, then recovered gently to 9 in 2023, 10 in 2024 and 10 in 2025 — about 10 a year, or roughly 470–720 beds.

Source: MetroEngines Research & Consulting (OTA-listing-confirmed basis, N=224 properties opened 2015–2026). Because OTA listing can precede or follow the opening, the most recent years may increase with future listings

For 2026, 6 openings with 375 beds have been confirmed as of October, but this figure is not final. Because OTA listing can precede or follow the opening, recent openings may not have been fully captured at the time of tabulation, and the 2026 count should be read on the assumption that it will rise over time. Looking individually at openings over the last two years, three types are growing in parallel: station-front and terminal locations in regional cities (Matsue, Fukui, Nagasaki, Beppu), airports (Naha Airport), and urban sauna-focused formats (Azabu-Juban, Shinagawa, Akasaka). The distribution is shifting away from the Tokyo-centric pattern of the late 2010s.

Source: MetroEngines Research & Consulting (OTA-listing-confirmed basis. Opening months are based on listing information. 2026 as of October)

We also check construction-start trends for the lodging industry as a whole, regardless of type. According to e-Stat “Building Construction Starts Statistics”, nationwide starts of lodging-use buildings fell from 2,300 in 2019 to 1,422 in 2021, then returned to 2,374 in 2022, 2,222 in 2023 and 2,836 in 2024, above the 2019 level. Floor area started, however, peaked at about 3.04 million ㎡ in 2018 and was about 1.76 million ㎡ in 2024, and floor area per building more than halved from about 1,435 ㎡ in 2018 to about 619 ㎡ in 2024. More buildings, each smaller: starts of small lodging facilities are lifting the total. In Tokyo, 67 buildings with about 530,000 ㎡ were started in 2024, with large projects contributing heavily to floor area. Lodging-focused supply, including capsule hotels, sits within this “many buildings, small floor area” trend.

We also note the demand-side environment. In the Japan Tourism Agency’s Overnight Travel Statistics Survey (first preliminary figures for August 2026), total guest nights in August were 62.68 million, down −6.7% YoY, of which foreign guest nights were 12.64 million, down −10.3%, an adjustment from the previous year’s high level. Note that from January 2026 the survey’s stratification basis changed from “number of employees” to “number of rooms,” so YoY comparisons include the effect of this revision. Read in this context, this article’s observation — that capsules filled while raising rates in a month when guest nights were below the prior year — carries more weight.

Which Cities and Price Bands Still Have Headroom — Cities with 2–3 Capsule Beds per 100 Hotel Rooms

To compare supply depth across cities, the ratio to existing hotel rooms is more appropriate than raw property counts. Calculating capsule beds per 100 rooms of combined business, city and deluxe hotels in each prefecture gives 9.2 beds in Tokyo, 6.4 in Kyoto and Kanagawa and 6.0 in Osaka — high density in the centers of the Tokyo metropolitan area and Kansai. By contrast, Fukuoka is at 3.3, Chiba 3.0, Hyogo 2.8, Aichi and Okinawa 2.6, Hokkaido 2.2 and Shizuoka 1.1; even in regions with ordinance-designated cities, capsule density is around one-third of Tokyo’s or less.

Overlaying this density with sell-through reveals where headroom lies. Fukuoka (density 3.3, sell-through 98.9%, N=5), Hokkaido (2.2, 98.1%, N=4), Aichi (2.6, 98.4%, N=3) and Chiba (3.0, 98.1%, N=3) all have low supply density, and their existing properties — though few — exceed Tokyo’s sell-through rate. Repeating that these are indicative small-sample figures, these four regions are where existing capsule supply is thin and still filling up. Tokyo, conversely, maintains 96.3% despite having the highest density, meaning demand is absorbing the depth of supply. Kyoto has a density of 6.4 against sell-through of 94.8% (N=3), the lowest among the 11 regions, and its all-property rate of 83.9% shows looser August supply and demand than elsewhere.

Source: MetroEngines Research; compiled by HotelBank Editorial Team (property and bed counts are confirmed-operating basis as of October 2026. Hotel rooms are the total of business, city and deluxe hotels. Sell-through rate is for August 2026. Indicative values omitted for regions with capsule N=2)

On price bands, the starting point is that capsule listed prices (2 guests) — ¥11,900 in Tokyo, ¥11,800 in Osaka, ¥10,200 in Fukuoka and ¥8,600 in Kanagawa — are 60–70% of business hotel levels in the same prefecture. On estimated settled ADR, capsules are ¥5,400 per bed in Tokyo, ¥5,700 in Fukuoka and ¥4,000 in Osaka, a gap of ¥5,000–¥8,000 versus a single room at a business hotel. The larger this gap in a city, the more easily capsules function as a receptacle for demand pushed out by rising hotel rates. This view is consistent with a capsule bed manufacturer citing growing inbound travel and soaring lodging rates as drivers of capsule demand in its August 2026 new product announcement.

For operators, the implications are as follows. First, in regions where sell-through exceeds 96%, there is room to widen listed inventory gradually. A state where remaining OTA availability touches zero for most of the month may mean missed opportunities to test pricing; increasing listed inventory and then differentiating prices by demand day could be an option to grow total revenue. Second, from an opening or investment perspective, terminal areas in Fukuoka, Nagoya, Sapporo and Chiba — where density is low and existing properties are full — are candidates. Third, value-adds such as saunas, large public baths and women-only floors appear in public plan names (large bath in 12%, women-oriented in 21%), and are factors supporting a move from rates in the ¥5,000s to the ¥7,000s. All of these are directional findings based on simple aggregation, and individual investment decisions require location-specific demand studies. The land price ceiling at which a capsule format works and floor efficiency per bed are back-calculated in Capsule Hotels Can Carry ¥8.99M/Tsubo Land, 1.28x Business Hotels.

Summary

In August 2026, Tokyo’s 27 capsule hotels posted a 96.3% sell-through rate, the highest of eight property types, and capsules beat the all-property average in all 11 prefectures with three or more. However, capsules have a low median OTA listing ratio of 13%, and restricting to listing ratios of 30% or more leaves not a single capsule hotel, so 96% should be read not as actual occupancy but as an upper bound showing that “thinly listed inventory repeatedly sold.” Even so, the facts that the daily low never fell below 94%, that capsules ranked first for three straight months, and that their estimated settled ADR rose +4.6% in a month when market-wide ADR fell −9.2% point to supply-demand strength that surface effects cannot explain.

Supply totals 237 properties and 30,453 beds nationwide, with 41% in Tokyo and 69% in four prefectures. After the 2017–19 wave, openings have run at about 10 a year, and in 2024–26 they are spreading across three types: regional terminals, airports and urban sauna formats. Fukuoka, Hokkaido, Aichi and Chiba, where capsule density is a low 2–3 beds per 100 hotel rooms and existing properties fill at 98%+, still have supply headroom — with the caveat that these are indicative small-sample figures. The capsule format is the receptacle where supply and demand tighten most when hotel rates keep rising, and its supply headroom is expanding beyond Tokyo.

Frequently Asked Questions