According to the Japan Tourism Agency’s (JTA) “Overnight Travel Statistics Survey, July 2026 — First Preliminary Report,” released on August 31, 2026, total guest-nights in July came to 54.67 million, down 3.5% year on year (YoY), of which foreign guests accounted for 13.58 million, down 4.6%. Deriving the Japanese share from the difference gives 41.09 million guest-nights, down 3.1%. The decline narrowed from June (-6.3%), but the growth in Japanese overnight stays that regional Japan enjoyed through the first half of 2026 reversed in July on a single-month basis. Meanwhile, aggregating MetroEngines Research’s estimated settled ADR across all 47 prefectures shows that the July room rate fell 4.4% on a national weighted-average basis (17,060 properties) — in other words, price adjusted ahead of volume. In this article we match the YoY change in guest-nights against the YoY change in estimated settled ADR for each of the 47 prefectures, classify them into four volume × rate quadrants, and set out the levers for prefectures with the most room for volume recovery and those holding up revenue through rate.

Reading the July First Preliminary Report — deriving the Japanese figure from the difference

The First Preliminary Report showed total guest-nights of 54.67 million (-3.5%) and foreign guest-nights of 13.58 million (-4.6%). The Japanese figure can be derived from the difference: “Japanese guest-nights = total − foreign = 54.67M − 13.58M = 41.09M.” The prior-year figures can be back-calculated as “total 54.67M ÷ (1 − 0.035) = 56.653M” and “foreign 13.58M ÷ (1 − 0.046) = 14.235M,” giving “Japanese 56.653M − 14.235M = 42.418M,” and the YoY change is “41.09 ÷ 42.418 − 1 = -3.1%.” The official Japanese figure published by the JTA in the same release is 41.10 million guest-nights, -3.1%; the 10,000 guest-night difference can be attributed to rounding to units of 10,000.

The key point here is that although the foreign decline (-4.6%) was steeper than the Japanese decline (-3.1%), in absolute guest-nights the Japanese decrease (about 1.33 million) was twice the foreign decrease (about 0.66 million). Because Japanese guests account for 75% of the total, a -3% move among Japanese guests has a market impact comparable to a -5% move among foreign guests. The Japanese decline halved from -6.3% in June to -3.1% in July, but the fact that the “regional growth of the first half” broke off in July is significant.

Note that the Second Preliminary Report, released on September 30, revised the July figures downward: total 53.97 million guest-nights (-4.7%), Japanese 40.66 million (-4.1%), and foreign 13.31 million (-6.5%). The valid response rate was 32.9% for the First Preliminary Report and 45.1% for the Second; because responses from smaller properties increase as collection progresses, preliminary figures are routinely revised. We measured how much preliminary figures move over the past 13 months in How Much Are Overnight Travel Statistics Preliminary Figures Revised? — Up to 3.4% for Guest-Nights, Within 1.0pt for Occupancy. The prefecture-level volume figures in this article use the Second Preliminary Report, which includes a prefectural breakdown.

Source: Compiled by HotelBank Editorial Team from the Japan Tourism Agency’s “Overnight Travel Statistics Survey,” July 2026 First Preliminary Report (released Aug 31) and Second Preliminary Report (released Sep 30). Japanese figure by difference calculated by the editorial team.

The JTA notes that from the January 2026 survey it changed the stratification criterion from “number of employees” to “number of guest rooms,” and that YoY comparisons may include the effects of this revision. Prefectures showing swings of more than ±20% (Ehime +36.6%, Kochi +28.0%, Yamanashi -21.3%, etc.) should be read on the assumption that both changes in demand and the change in estimation method are at play. We explain how to read this survey in What Is the Overnight Travel Statistics Survey? How to Read the JTA’s Accommodation Statistics.

Estimated settled ADR down 4.4% nationally — still -3.2% excluding Osaka; business hotels the most resilient by type

So how did rates move? Aggregating MetroEngines Research’s estimated settled ADR (excluding tax) across the 47 prefectures, the national weighted average for July 2026 was approx. ¥10,100, down 4.4% from approx. ¥10,600 a year earlier (17,060 properties). The largest drag was Osaka at -30.7% (approx. ¥8,800, 647 properties): July 2025 fell during Expo 2025 Osaka, Kansai, so the prior-year rate level itself was elevated. Re-aggregating for the 46 prefectures excluding Osaka still leaves a YoY decline of -3.2%, but the drop narrows by just over one point.

By region, the weighted average for the eight prefectures of the three major metropolitan areas (Saitama, Chiba, Tokyo, Kanagawa, Aichi, Kyoto, Osaka, Hyogo) was -9.7%, versus -2.0% for the other 39 prefectures — the rate adjustment was concentrated in urban areas. The JTA’s Second Preliminary Report points the same way, with foreign guest-nights down 9.2% in the three major metropolitan areas and down 0.9% in regional areas; in the cities, “foreign volume” and “rate” adjusted at the same time.

Source: Compiled by HotelBank Editorial Team from MetroEngines Research (July 2026 estimated settled ADR, all property types; see the tables in the text for the number of properties per prefecture)

Only 10 of the 47 prefectures posted a positive YoY ADR change. The top five were Akita +8.9%, Tokushima +8.0%, Okinawa +7.1%, Shiga +3.3% and Kochi +3.1%; the bottom five were Osaka -30.7%, Nara -13.3%, Kyoto -12.8%, Yamaguchi -10.1% and Kumamoto -10.0%. Nara and Kyoto can be read as part of a Kansai-wide move linked to Osaka’s price adjustment — guest-nights in Nara actually rose 11.5%.

Source: Compiled by HotelBank Editorial Team from MetroEngines Research (ADR, number of properties) and the Japan Tourism Agency’s “Overnight Travel Statistics Survey,” July 2026 Second Preliminary Report (guest-nights, room occupancy rate)

By property type, the order of resilience was business hotels, resorts, ryokan, then city hotels. Business hotels could be aggregated with 20 or more properties in all 47 prefectures; 30 of them posted YoY gains, and the weighted average of -3.7% (-0.6% excluding Osaka) was nearly flat. Ryokan rose in 13 of 46 prefectures with a weighted average of -4.9%; resorts rose in 7 of 18 prefectures at -1.5%; and city hotels rose in only 2 of 15 prefectures at -12.1% — the higher the urban price tier, the larger the adjustment. Overlaying the JTA occupancy data — ryokan +3.0pt and resorts +2.1pt above last year, versus business -0.1pt and city +0.3pt, essentially flat — July was a month in which “ryokan and resorts went after occupancy while urban city hotels gave up rate.”

Source: Compiled by HotelBank Editorial Team from MetroEngines Research (estimated settled ADR, excl. tax; each cell is the area median; N is the number of properties covered in July 2026)

Volume × rate quadrants — half of the 47 prefectures landed in “volume down, rate down”

To sum up so far, July was below last year on both volume and rate at the national average level, but the combination varies widely by prefecture. We therefore plotted YoY change in guest-nights (Second Preliminary Report, total) on the horizontal axis and YoY change in estimated settled ADR on the vertical axis, dividing the 47 prefectures into four quadrants.

Source: Compiled by HotelBank Editorial Team from the Japan Tourism Agency’s “Overnight Travel Statistics Survey,” July 2026 Second Preliminary Report (horizontal axis) and MetroEngines Research (vertical axis, estimated settled ADR)

The result is clear. “Volume down, rate down” is the largest group with 19 prefectures, accounting for 50.7% of guest-nights. That is because the major cities all landed here — Tokyo (volume -10.7% / rate -4.8%), Osaka (-14.9% / -30.7%), Kanagawa (-12.2% / -7.9%), Kyoto (-6.8% / -12.8%) and Aichi (-10.9% / -3.3%) — along with prefectures dependent on short-distance leisure trips from the Tokyo and Nagoya areas, such as Yamanashi (-21.3% / -6.4%) and Gifu (-8.9% / -4.7%).

In contrast, “volume up, rate down” comprises 17 prefectures with a 24.3% share, lined with prefectures whose occupancy also beat last year, such as Ibaraki (volume +23.3% / rate -0.6%), Wakayama (+19.8% / -3.0%), Toyama (+19.7% / -0.3%), Kagawa (+14.3% / -8.0%) and Saitama (+10.2% / -3.1%). A situation where volume is recovering but rates remain below last year means, conversely, that there is room to recover rates step by step.

“Volume down, rate up” covers 6 prefectures with a 21.2% share: Okinawa (volume -10.3% / rate +7.1%), Hokkaido (-9.6% / +1.4%), Tokushima (-10.9% / +8.0%), Yamagata, Nagano, and Fukuoka, where the rate was essentially flat (-0.0%). Okinawa and Hokkaido maintained rates despite near double-digit declines in Japanese guest-nights (-13.3% and -7.8%), making them prefectures that supported July revenue through rate. Finally, “volume up, rate up” is limited to five prefectures with a 3.9% share: Ehime (+36.6% / +1.9%), Kochi (+28.0% / +3.1%), Akita (+18.9% / +8.9%), Aomori (+17.1% / +2.2%) and Shiga (+9.0% / +3.3%). A defining feature of July was that three of Shikoku’s four prefectures landed on the volume-up side, while Akita and Aomori in northern Tohoku secured gains in both volume and rate.

Source: Compiled by HotelBank Editorial Team from the Japan Tourism Agency’s “Overnight Travel Statistics Survey,” July 2026 Second Preliminary Report, and MetroEngines Research. Quadrant boundaries are 0% YoY. Fukuoka’s ADR was -0.0% (¥10,873 → ¥10,874) and is classified on the rate-maintained side.

Splitting Japanese and foreign guests reveals what lies behind the “volume down” in the major cities. Tokyo saw Japanese -14.8% versus foreign -7.0%, Osaka Japanese -18.1% versus foreign -10.5%, and Okinawa Japanese -13.3% versus foreign -2.2% — in every case the Japanese decline was larger. JTB’s outlook for the summer holidays (July 15 to August 31) projected 69 million domestic travelers (95.6% of the prior year) and spending of ¥48,500 per person (103.2%), analyzing that a thrift mindset would outweigh rising wages and bonuses. If Japanese travelers “take fewer trips while maintaining spending per trip,” volume in the major cities adjusts first and rates move later. The July combination in Tokyo and Osaka is consistent with this view.

Source: Compiled by HotelBank Editorial Team from the Japan Tourism Agency’s “Overnight Travel Statistics Survey,” July 2026 Second Preliminary Report

As for foreign guests, the Japan National Tourism Organization (JNTO) reported 3,442,100 international visitor arrivals in July, up 0.1% YoY and a record for the month of July, yet foreign guest-nights fell 6.5%. Guest-nights can fall even when arrivals are flat because the -56.1% drop in the China market was offset by short-haul markets — South Korea +31.9%, Taiwan +26.4%, Hong Kong +54.8% — shifting the mix toward visitors with shorter stays. Fukuoka’s +18.0% growth in foreign guest-nights likely reflects its role as a destination for this short-haul shift. The weight of the China decline in overall foreign guest-nights, and the prefectural changes excluding China, are analyzed in detail in July Foreign Guest-Nights −4.6%, 93% of the Decline from China — −1.0% Excluding China, with 18 Regional Prefectures Growing.

Accommodation CPI +2.0% vs. estimated settled ADR -4.4% — the gap between the statistical price index and market rates

How does the YoY decline in rates show up in the official price index? In the Ministry of Internal Affairs and Communications’ Consumer Price Index (2020 base), the item “accommodation charges” stood at 169.1 in July 2026, up 2.0% YoY. Estimated settled ADR for the same month was -4.4%, a gap of 6.4 points.

Source: Compiled by HotelBank Editorial Team from the Statistics Bureau of Japan’s “Consumer Price Index” (statistical table ID 0003427113, item “accommodation charges,” national, 2020 base) and MetroEngines Research (estimated settled ADR, 47-prefecture weighted average)

The direction of the gap has been consistent. After both stood level at +8.4% in October 2025, YoY ADR outpaced CPI from November through January 2026 (in November, ADR +11.9% vs. CPI +9.3%); from April 2026 ADR fell below CPI and turned negative at -4.6% in June, with CPI following about two months later at +2.0% in July and -0.2% in August. Accommodation CPI is an index that continuously tracks prices under specific conditions at surveyed properties, whereas estimated settled ADR takes as its population the lowest-plan levels of all properties with confirmed operations, so it directly reflects the weight of areas such as Osaka where the prior-year price level was exceptional. Market rates swing more widely than the index and turn earlier, so for monthly operating decisions it is the market-rate series, not the index, that operators should watch (for how far back accommodation CPI and prefectural ADR can be traced, see How to Look Up Past Hotel Prices — Accommodation CPI from 1998, Prefectural Data from July 2019).

Source: Compiled by HotelBank Editorial Team from the Statistics Bureau of Japan’s “Consumer Price Index” and MetroEngines Research. National ADR is the weighted average of the 47 prefectural medians, weighted by number of covered properties.

That said, the fact that CPI turned negative in August 2026 at -0.2% YoY should not be overlooked. It is the first time accommodation CPI has fallen below the prior year in 3 years and 5 months, since March 2023 (-0.6%), when the effects of the nationwide travel subsidy program were still present — suggesting that the adjustment in market ADR has begun to appear in official statistics with a two-month lag. The JTA’s August First Preliminary Report showed 62.68 million guest-nights, down 6.7% (Japanese -5.8%, foreign -10.3%), a deeper volume adjustment than in July. We plan to revisit August rates alongside the Second Preliminary Report (scheduled for release on October 30).

Levers for operators — prefectures with the most room for volume recovery, and those supported by rate

The four-quadrant classification translates directly into operator priorities. We go through each quadrant in turn.

Volume down, rate down (19 prefectures including Tokyo, Osaka, Kanagawa, Kyoto, Aichi and Yamanashi) — the most room for volume recovery. Occupancy in this quadrant remains high — Tokyo at 75.3% (No. 1 nationally, +1.8pt) and Osaka at 72.6% (No. 2) — so the volume decline reflects a high prior-year base and Japanese thriftiness rather than an occupancy ceiling. For Osaka, which is being compared against last year’s Expo period, the rate picture will change from October as the YoY comparison switches to a post-Expo base. As levers, building out sales channels to capture growth from short-haul markets (South Korea, Taiwan, Hong Kong) and expanding multi-night and value-added plans suited to Japanese travelers’ “fewer trips, same spend per trip” behavior are the starting points for recovering volume. With rates having bottomed out while occupancy remains high, there is room to restore them step by step as demand returns. For rates in the Kansai region, see also Sakai Station Hotel Rooms Grew 1.4x to 1,715 in 2025 — Post-Expo ADR of ¥7,449.

Volume up, rate down (17 prefectures including Ibaraki, Wakayama, Toyama, Kagawa, Nara and Saitama) — time to recover rates. Prefectures where rates are below last year despite double-digit growth in guest-nights are at the stage before demand recovery has been reflected in pricing. In prefectures where occupancy has clearly risen — Ibaraki (+4.2pt), Wakayama (+5.2pt), Nara (+6.1pt), Toyama (+4.1pt) — revisiting the lowest-plan price level starting with high-demand days of the week and periods is the shortcut to rate recovery. This quadrant overlaps heavily with prefectures that grew Japanese overnight stays in the first half, and that first-half momentum continued on the volume side in July.

Volume down, rate up (Okinawa, Hokkaido, Tokushima, Yamagata, Nagano, Fukuoka) — prefectures supported by rate. Okinawa posted rate +7.1% despite Japanese guests -13.3%, and Hokkaido rate +1.4% with Japanese guests -7.8%, protecting revenue through rate. With occupancy at 61.8% (-2.6pt) in Okinawa and 69.4% (-0.4pt) in Hokkaido, both have more headroom than Tokyo or Osaka, so the next move is volume recovery. In Okinawa, foreign guests were nearly flat at -2.2%, so the focus is how to bring back Japanese summer demand. Revisiting price bands for weekdays and shoulder periods, and designing multi-night stays that extend length of stay while holding rate, are options for recovering volume without eroding rate.

Volume up, rate up (Ehime, Kochi, Akita, Aomori, Shiga) — July’s growth prefectures. Shikoku and northern Tohoku secured gains in both volume and rate. Kochi, which we noted in our first-half article as a prefecture where occupancy exceeds pre-COVID levels, kept growing in July at +28.0%. What to keep in mind in this quadrant is that the YoY figures may include the effect of the change in estimation method, so operators should confirm that the same trend continues in subsequent months before moving on to investment decisions on facilities and staffing.

Looking at the whole picture, July’s volume adjustment occurred in the major cities and centered on Japanese guests, while the rate adjustment, including Osaka’s base effect, was concentrated in urban city hotels. Regional ADR was -2.0%, and business hotels were -0.6% excluding Osaka — not as weak as the national average suggests. The momentum of the “prefectures that grew on Japanese demand” in the first half continues on the volume side in many prefectures, and July’s single-month reversal is best read mainly as a simultaneous adjustment of volume and rate in the major cities.

Frequently Asked Questions

Q. Should I use the First or the Second Preliminary Report?
A. In practice, use the First Preliminary Report to gauge the market right after release, and the Second Preliminary Report — which includes breakdown tables — for comparisons by prefecture or facility type. This article likewise starts from the First Preliminary Report for national figures and uses the Second Preliminary Report for prefectural analysis.

Q. Why does estimated settled ADR move in a different direction from accommodation CPI?
A. Because the population and aggregation method differ. CPI continuously tracks prices under specific conditions at surveyed properties, while estimated settled ADR is estimated from the lowest-plan levels of all properties with confirmed operations. The market-rate series tends to swing more widely and turn earlier; in 2026, ADR turned negative YoY in June and CPI in August.

Q. How should Osaka’s ADR of -30.7% be interpreted?
A. It is largely a base effect: July 2025 fell during Expo 2025 Osaka, Kansai, so the prior-year rate level was exceptional. Occupancy remains at 72.6%, second-highest in Japan. From October the comparison switches to a post-Expo base, so the YoY picture will change.

Related Articles

• Prefectures Where Japanese Overnight Stays Grew in H1 2026 — Tottori and Ibaraki +27.7%, Kochi Occupancy +7.2pt Above Pre-COVID
• What Is the Overnight Travel Statistics Survey? How to Read the JTA’s Accommodation Statistics — Occupancy 61.6%, Simple Lodgings 29.6%, Preliminary vs. Final [2025]
• How to Look Up Past Hotel Rates — Using e-Stat and Reading Monthly ADR, with 24 Months of Kanazawa Data
• Mitsubishi Estate’s 10,000 sqm Hotel in Dotonbori — Required ADR of ¥19,000–34,000, 1km Radius Supply Up 2.1x in 10 Years

* From the January 2026 survey, the JTA’s “Overnight Travel Statistics Survey” changed its stratification criterion from “number of employees” to “number of guest rooms,” so YoY changes and differences may include the effects of this revision. * August 2026 guest-nights are First Preliminary figures and may be revised in the Second Preliminary figures scheduled for release on October 30, 2026. * ADR in this article is an estimate (median error of about 7% against REIT-disclosed results); settled ADR disclosed by REITs is shown excluding consumption tax, and the estimated settled ADR in this article is likewise on an excluding-tax basis.