“Autumn in Japan draws inbound visitors chasing the fall colors” — trace that industry commonplace across twelve months of official prefecture-level statistics and a rather different picture emerges. Calculating each prefecture’s “foreign guest nights ÷ total guest nights” on a monthly basis from the Japan Tourism Agency’s Accommodation Survey, only 6 of the 47 prefectures see their foreign guest share peak in autumn (September–November). The largest group is winter-peak, with 18 prefectures, followed by spring-peak with 17. This article redraws that seasonal map of foreign guest share, then tests whether estimated transacted ADR and inventory sell-through move in the same direction in the prefectures where the share does rise in autumn.

The national foreign guest share is 26.7% — but inbound demand itself has almost no seasonality

Start with the national picture. Across the twelve months from June 2025 to May 2026, total guest nights nationwide came to 650.82 million, of which 173.90 million were foreign guest nights — a share of 26.7%. Month by month, the low was 21.0% in August 2025 and the high 31.3% in April 2026, a swing of 10.3 points.

What matters here is what creates that swing. Convert foreign guest nights themselves into a seasonal index and the result is almost perfectly flat: summer 0.98, autumn 1.02, winter 0.98, spring 1.02. Japanese guest nights, by contrast, tilt clearly toward summer and sag in winter — summer 1.10, autumn 1.06, winter 0.88, spring 0.95. In other words, the “seasonal variation in foreign guest share” observed at the national level is largely an apparent movement: inbound visitors are not rising and falling with the seasons; the seasonality of Japanese demand is moving the denominator. Foreign share drops in summer not because inbound visitors decline, but because Japanese guests increase. For the structure of the receiving side — which prefectures inbound demand lands in, and from which nationalities — see our Prefecture Host Map 2026 for inbound guest nights by nationality, which breaks it down prefecture by prefecture.

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey time-series tables

Sort the 47 prefectures into four types and only 6 are autumn-peak

Running the same calculation by prefecture, we derived seasonal indices for foreign guest share across four seasons — summer (June–August), autumn (September–November), winter (December–February) and spring (March–May) — and typed each prefecture by the season with the highest index. Prefectures whose index range across the four seasons is under 0.15 are treated as having weak seasonality and assigned to a “flat” type.

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey time-series tables

Autumn-peak accounts for just 6 prefectures, or 13%. Fall foliage may be widely recognized as one of Japan’s tourism assets, but the prefectures whose foreign guest nights are structurally concentrated in autumn are few. In sheer numbers, winter and spring split the market between them — a reflection, it seems reasonable to conclude, of two already established inbound seasons: ski demand and cherry blossom demand.

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey time-series tables

Inside the 6 autumn-peak prefectures — the share rises because inbound visitors really do increase

There is one question worth settling here. When share rises in autumn at the prefecture level, is it because inbound visitors are increasing, or merely because Japanese guests are declining? Nationally the latter effect dominated. What about in these 6 prefectures?

The answer is unambiguous: in all six, foreign guest nights themselves increase in autumn. The autumn index for foreign guest nights is 1.35 in Wakayama, 1.30 in Tochigi, 1.27 in Kagawa, 1.26 in Ibaraki, 1.18 in Mie and 1.10 in Shizuoka. The corresponding autumn index for Japanese guest nights runs 0.93–1.08 — essentially flat in every prefecture. Inbound demand, flat at the national level, tilts distinctly toward autumn in these six. The rise in share comes from an expanding numerator, not a shrinking denominator.

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey time-series tables

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey time-series tables

Separating structure from event — only 3 prefectures were autumn-type two years running

Declaring a prefecture “autumn-type” on a single year of data is risky. So we applied the same classification to the prior-year period (June 2024–May 2025). Three prefectures were autumn-peak for a second consecutive year: Ibaraki (prior-year autumn index 1.20), Wakayama (1.28) and Tochigi (1.24). The remaining three — Kagawa, Mie and Shizuoka — were spring-peak in the prior year and swung to autumn over the past twelve months.

Kagawa’s shift has a clear cause. The autumn session of the Setouchi Triennale 2025 ran for 38 days from October 3 to November 9, 2025, with 216 artists and groups from 37 countries and regions taking part. The prefecture’s foreign guest share hit 29.7% in October 2025 and 27.2% in November, standing out for the year and more than doubling the 13.4% recorded in the immediately preceding August. A one-off factor — an international art festival held once every three years — effectively rewrote the seasonal shape of the share itself. There is no guarantee the same shape recurs in autumn 2026. For which price tiers within the prefecture the festival lifted, our ADR tier map of Kagawa’s 16 municipalities breaks it down by municipality and price tier.

Ibaraki, Wakayama and Tochigi, by contrast, have traced the same shape two years in a row. In Ibaraki, the kochia at Hitachi Seaside Park turn crimson and reach their peak in mid-October; Wakayama has the autumn of Mount Koya and Kumano; Tochigi has the fall colors of Nikko. All three hold landscape assets fixed to a season. Because this is demand tied to the calendar rather than dependent on an event, the likelihood of the same shape reappearing next autumn is comparatively high.

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey time-series tables

In the autumn when foreign share rises, rates are not rising

So in autumn, when foreign share reaches its annual high, do room rates move in the same direction? Using MetroEngines Research’s estimated transacted ADR (a monthly series on a verified-history basis), we calculated seasonal indices over the same 12-month window as the foreign guest share.

The result is essentially “no movement.” The autumn ADR seasonal index is 0.99 in Ibaraki, 0.97 in Wakayama, 0.98 in Kagawa, 1.06 in Tochigi, 0.94 in Mie and 0.97 in Shizuoka. Only Tochigi came in above its annual average; in the other five, autumn ADR sat slightly below the annual average. That is a striking flatness against foreign share indices that jump to 1.09–1.33. The fact that price swings remain at the daily level even where monthly averages look flat is covered in our ranking of daily price ranges for September–October 2026.

Source: compiled by the HotelBank Editorial Team from the Japan Tourism Agency’s Accommodation Survey / MetroEngines Research

This divergence is, in a sense, the natural consequence of rates being set by total demand. Annual foreign share in the six prefectures is in single digits for many — Ibaraki 4.6%, Tochigi 4.6%, Mie 4.0%, Shizuoka 9.1% — and even in Wakayama at 17.4% and Kagawa at 21.8%, domestic guests still account for roughly 80% of demand volume. Even if inbound visitors rise 30% in autumn, the increment as a share of total accommodation demand stays within a few points, which is unlikely to exert enough pressure to lift rates. Rather, it is the two domestic peak periods — summer holidays and year-end — that create the rate peaks: Shizuoka’s August ADR of ¥16,600 was its annual high, and Mie’s December figure of ¥13,900 was its highest.

Estimated transacted ADR (¥, tax-exclusive equivalent). Median covered property counts: Ibaraki 308, Wakayama 222, Kagawa 154, Tochigi 395, Mie 357, Shizuoka 827. All months on a verified-history basis. Source: compiled by the HotelBank Editorial Team from MetroEngines Research

Rates not moving does not mean there is no revenue opportunity. If anything the reverse: if the price structure tracks the annual average through a period when the composition of demand is changing, then upside remains in reworking the pricing design. One reading is that there is room to align value-adding elements — in-house signage, payment methods, language support in food and beverage service — with the month when the inbound share is at its annual high, and to reflect that incrementally in pricing.

Which property types move autumn inventory first — a completed observation month and a forward snapshot

Finally, we look at supply and demand from the inventory side. There is a clear data constraint here. Our OTA inventory data collection began on March 10, 2026, so no inventory history exists for autumn 2025. Autumn 2026, meanwhile, is still mid-observation, and the further back the lead time, the fewer properties are consistently covered. This article therefore limits itself to two views: (1) occupancy by property type for the most recent month for which observation is complete, and (2) an inventory sell-through cross-section as of the survey date for Saturday, October 3, 2026, a representative autumn weekend, by which point observed properties were nearly all in place. We do not compare sell-through speed by lead time, because the observed population is not consistent.

Occupancy for July 2026 (completed observation month), estimated on an OTA-listed inventory basis; covered property counts in parentheses. Source: compiled by the HotelBank Editorial Team from MetroEngines Research

As of July 2026, in the middle of the summer peak, all six prefectures were pinned at a high 82.7–85.9% on an all-property basis, with little variation between property types. The highest within any prefecture was Ibaraki’s resort category at 92.1% (15 properties); the lowest was Kagawa’s ryokan category at 74.2% (31 properties). This period is driven by domestic demand and falls in a season when the foreign share is low.

Next we look at the inventory sell-through cross-section as of the survey date for a weekend in October, when the foreign share rises. Saturday, October 3, 2026 sits at LT57 — that is, 57 days before check-in — and at that point more than 90% of observed properties were included in the data.

Source: compiled by the HotelBank Editorial Team from MetroEngines Research

By property type, city hotels showed the most advanced sell-through in all six prefectures (in Wakayama alone, roughly level with business hotels). Tochigi’s city hotels had 71 rooms remaining out of 1,487 (95.2% occupancy, 14 properties, as of LT57 for October 3, 2026), and Ibaraki’s city hotels 97 rooms remaining out of 1,823 (94.7%, 16 properties) — in both cases most of the inventory had moved a full two months before check-in. Ryokan, by contrast, still hold room to spare: 71.7% in Tochigi (1,978 of 6,987 rooms remaining, 177 properties observed) and 62.6% in Shizuoka (3,647 of 9,761 rooms, 386 properties observed).

This gap is not simply a matter of “city hotels being popular”; it reflects differences in the scale of room inventory and in sales design. City hotels are a minority within each prefecture’s room mix (1,823 rooms in Ibaraki, 1,487 in Tochigi), and a smaller base makes sell-through look faster. Ryokan, meanwhile, have large room counts and commonly operate on the premise of holding inventory back to capture last-minute demand. The observation that “city hotels move first” is therefore best treated not as evidence that autumn inbound demand is concentrated in city hotels, but as a reflection of how differently each property type releases inventory and designs its lead times.

How many points does the autumn inbound upswing add to total demand?

We want to verify the point made above — that even if inbound visitors rise 30% in autumn, the increment as a share of total accommodation demand stays within a few points — using only figures already in this article. Rearranging the definition of foreign guest share (foreign guest nights ÷ total guest nights), the lift to total demand when Japanese guest nights are held at roughly their annual average can be written as annual foreign share × (autumn index for foreign guest nights − 1). No new observations are used; this is a unit conversion that simply substitutes the annual share and autumn index already shown in Table 2. That the autumn index for Japanese guest nights is flat at 0.93–1.08 across the six prefectures — confirmed above — is the premise behind this approximation.

In the four prefectures with single-digit annual share, even swinging the autumn index to the observed ceiling of 1.35 fails to produce a lift of 3.2 points. At each prefecture’s observed autumn index, the lift is 1.2 points in Ibaraki, 1.4 points in Tochigi, 0.7 points in Mie and 0.9 points in Shizuoka — all at or below 1.4 points. Even in Wakayama (6.1 points) and Kagawa (5.9 points), where annual share is a thick 17–22%, the figure stays in the high single digits. Seen against magnitudes of this size, the fact that the autumn ADR seasonal index fell below the annual average in 5 of the 6 prefectures is a coherent result.

Read across both axes, the lift ranges from a minimum of 0.4 points (share 4.0% × index 1.10) to a maximum of 7.6 points (share 21.8% × index 1.35) — a spread of nearly 19 times. What determines the magnitude is not the height of the autumn index so much as the thickness of the base, the annual share: at the same index of 1.35, a prefecture with 4.0% share gets 1.4 points while one with 21.8% share gets 7.6 points, a 5.5-fold difference. Put the other way, a prefecture with a thin share that wants to translate autumn inbound demand into rate cannot wait for total demand to lift it; it has to create a rate differential through product design aimed at inbound guests themselves. Note that this table restates ratios already published in units of demand volume, and is not a forecast of future demand or prices.

What this tells us

Three points are worth drawing together. First, the premise that “autumn equals inbound season” does not hold uniformly across the country. Foreign guest share peaks in autumn in only 6 of the 47 prefectures; the market’s mainstream is winter (18 prefectures) and spring (17). Launching an “autumn inbound initiative” without knowing which type your own catchment belongs to risks an investment misaligned with the actual shape of demand.

Second, in the 6 autumn-peak prefectures inbound visitors genuinely do increase. Inbound demand, flat nationally, moves clearly higher in these six at an autumn index of 1.10–1.35. But a single year of data cannot distinguish structure from event. Only three prefectures — Ibaraki, Wakayama and Tochigi — held the autumn pattern for a second consecutive year, and Kagawa’s shift into autumn contains a strong one-off factor in an international art festival held once every three years. When planning for next autumn, that distinction is the practical fork in the road.

Third, the seasonal peak in foreign share and the seasonal peak in rate do not coincide. In 5 of the 6 prefectures, autumn ADR sits below the annual average, and rates remain governed by the domestic demand calendar of summer holidays and year-end. A state in which price stays at the annual average during the period when the inbound share is at its annual high indicates that the room to reflect a change in the composition of demand in price remains untouched. Aligning in-house operations and product design with the autumn guest mix, then working that value into price step by step — for the autumn-peak prefectures, that is likely the most realistic direction for revenue growth.

Related reading

• US Now Tops Japan Inbound at 13.5% — Prefecture Host Map 2026
• Kagawa’s 16 Municipalities: 9.4x ADR Gap and Four Autumn Price Tiers
• Double Ten x Sports Day: Japan’s Oct 9-12 Capacity at 76 Days Out
• Nasu vs Morioka Lodging Tax Oct 1: Tiered 1.0-3.0% vs Flat ¥200
• Japan Autumn 2026 Hotel Price Range: 3.3x CV Gap Across 47 Prefectures
• Japan Hotel Industry Trends 2027: ADR +3.9%, Inbound Shift, New Supply