In the January visitor-arrival estimates published by the Japan National Tourism Organization (JNTO) on February 18, 2026, South Korea recorded 1,176,000 visitors (+21.6% YoY) — the first time any single market has exceeded 1.1 million in one month. Growth has not let up since: in the September estimate (published October 21), the latest available as of our research date (October 23, 2026), Korea logged 842,300 visitors (+25.6%), a record for September, bringing the January–September cumulative total to 8,262,700 (+21.6%). While total arrivals were running at −2.9% through September, Korea alone kept growing by more than 20% — and its annual peak falls in winter, above all in January. This article first confirms the monthly seasonality of the Korean market, then uses the Japan Tourism Agency’s Overnight Travel Statistics Survey for January 2026 (second preliminary release) to identify the prefectures where Korean guests stayed in January and their shares, and finally lines up how the estimated settled ADR of those host prefectures moved from October 2025 to January 2026 — organized so it can be used to design winter sales strategy.

Seasonality of the Korean Market — January Is the Annual Peak, +21.6% Through September

Let us start with the overall shape of the market. Overlaying JNTO’s estimates by nationality and month for 2024 through 2026 on the same January–December axis shows that, for three years running, Korea peaks in January, February and December and bottoms out in August–September (in 2026, June was the lowest month). In 2025, January had 967,100 visitors and December 974,239, while August was the year’s low at 660,917. January, February and December together accounted for 29.5% of the year (28.8% in 2024), well above the level the same calculation gives for total arrivals. For Korean travelers, Japan is a winter destination — and that pattern held again in 2026.

In 2026 the peak rose another notch. January reached 1,176,089 (+21.6% YoY) and February 1,086,381 (+28.2%), topping 1 million for two consecutive months. In its January release, JNTO cited the shift of the Lunar New Year from late January in 2025 to mid-February in 2026, along with stronger snow-season demand, as drivers of growth from Korea, Taiwan, Australia and other markets. In other words, January’s 1.176 million was built on underlying winter demand with no holiday effect, and February — which included the holiday — stacked further on top. The January–February total of 2,262,470 equals 27.4% of the January–September cumulative, above the same ratio in 2025 (26.7%) and 2024 (25.9%). The concentration in winter is intensifying.

The summer peak, by contrast, is low. August set a record for the month at 850,500 (+28.7%), and among the 14 markets that broke their August records, Korea had the most visitors of all 23 markets — yet in absolute terms it reached only a little over 70% of January, ranking sixth among 2026 months after January, February, May (951,328), July and April. September came in at 842,300 (+25.6%), a September record. JNTO cited athletes and officials arriving for the Asian Games in Aichi as a boost to the overall total, while explaining that many markets fell below the prior year due to higher airfares and reduced flights linked to the Middle East situation, and typhoon cancellations. That Korea sustained growth above 20% in this environment shows the resilience of this market.

Source: Compiled by HotelBank Editorial Team from JNTO, Visitor Arrivals by Nationality/Month (published October 21, 2026; 2026 figures provisional)

Source: Compiled by HotelBank Editorial Team from JNTO, Visitor Arrivals by Nationality/Month (published October 21, 2026). July–September 2026 are provisional figures; others are estimates

Where Korean Guests Stay in January — 67% of Volume in 4 Prefectures, Highest Shares in Kyushu, Shikoku and San’in

So where did January’s 1.176 million stay? According to the Japan Tourism Agency’s Overnight Travel Statistics Survey for January 2026 (second preliminary release), Korea ranked first in foreign guest nights by nationality (place of residence) with 2,047,670 guest nights (17.9% share, +6.4% YoY), narrowly ahead of Taiwan’s 2,041,930 (+3.8%). Third-placed China recorded 1,280,070 guest nights (−62.9%), the main reason total foreign guest nights for January fell 13.8%. Nationwide foreign guest nights were 12.83 million (−15.3%), and the Japan Tourism Agency notes that YoY comparisons may include the effect of the revised stratification criterion.

Ranking Korean guest nights by prefecture gives Tokyo 417,670, Osaka 387,710, Hokkaido 316,120 and Fukuoka 255,860 — these four alone account for 67.2% of Korean guest nights. Adding fifth-placed Okinawa (128,870) brings the share to 73.5%. In terms of volume, Korean winter stays are concentrated in the two big cities of Tokyo and Osaka, snow-season Hokkaido, and Fukuoka, the geographically closest destination.

The picture changes, however, when we look at Korea’s share of each prefecture’s foreign guest nights. In Kagoshima, Korea accounts for 41,880 of 60,760 foreign guest nights, or 68.9%; Ehime is at 65.9%, Yamaguchi 61.0%, Saga 49.8%, Oita 48.0%, Fukuoka 46.1%, Nagasaki 44.7% and Tottori 44.0% — 2.5 to 3.9 times the national average of 17.9%. Kyushu’s 7 prefectures combined host 421,150 Korean guest nights, only 20.6% of the Korean total, yet Korea’s share of Kyushu’s foreign guest nights reaches 46.6%. This contrasts with just 11.5% in Tokyo and 8.9% in Kyoto: Korean guests follow a two-tier distribution — “volume in the cities, dependence in western Japan’s onsen prefectures and Shikoku/San’in.”

Source: Compiled by HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey, January 2026 (second preliminary release), Reference Table 1 (properties with 20+ rooms)

Source: Compiled by HotelBank Editorial Team from the Japan Tourism Agency’s Overnight Travel Statistics Survey, January 2026 (second preliminary release), Reference Table 1 and foreign guest nights by prefecture. Nationality breakdown covers properties with 20+ rooms; prefecture totals cover all properties. YoY compares preliminary figures and may include the effect of the stratification change. ＊ = standard error rate of 30% or more

Looking at YoY for prefecture totals, Japan overall was −15.3%, Tokyo −13.0%, Osaka −24.5% and Hokkaido −18.6% — the big metro areas and snow country, where the drop in Chinese guests showed most, sank. Meanwhile many prefectures with high Korean shares exceeded the prior year: Oita +6.2%, Okinawa +12.6%, Yamaguchi +12.8%, Ehime +10.1%, Tottori +32.4%. Fukuoka (−8.1%) and Kagoshima (−5.2%) saw only modest declines. Allowing for the fact that these are preliminary-to-preliminary comparisons that include the stratification change, foreign stays in January 2026 held up better the higher a prefecture’s Korean share.

Winter Rates in Host Prefectures — Oita +18.1% and Saga +18.3% from October 2025 to January 2026

Next, we check how rates moved in winter in these prefectures using MetroEngines Research’s estimated settled ADR. We compare closed past months only, setting October 2025 — an autumn actuals month — against January 2026, Korea’s peak. Oita rose from ¥14,600 to ¥17,300, +18.1% (N=449 properties); Saga from ¥12,000 to ¥14,200, +18.3% (N=153); Kumamoto +13.6%, Ehime +14.3%, Kagoshima +11.7%, Kagawa +10.3%; and Fukuoka from ¥11,200 to ¥12,000, +7.1% (N=463). Every prefecture with a high Korean share was higher in winter than in autumn. Hokkaido (+6.5%) and Okinawa (+1.8%) also stayed positive.

Tokyo, Osaka and Kyoto stand in contrast. Osaka fell from ¥13,400 to ¥9,500 (−29.0%), Kyoto −31.4% and Tokyo −15.6% — but this mirrors the fact that October 2025 was the closing month of Expo 2025 Osaka, Kansai (closed October 13), which pushed autumn rates in Kansai and the Tokyo area to unusually high levels. In fact, January-over-January YoY remained positive — Osaka +5.5%, Tokyo +9.7%, Kyoto +1.1% — so winter rates did not collapse. Since the question here is “which prefectures lift rates in winter,” it makes sense to focus on host prefectures where autumn event effects are small.

What is interesting is the difference from the prior year. Running the same autumn→winter comparison one year earlier (October 2024 → January 2025), winter was lower in many host prefectures: Fukuoka −6.9%, Saga −10.7%, Nagasaki −7.9%, Hokkaido −5.6%, Kagoshima −1.8%. Only Oita (+3.3%), Nagano (+6.6%) and Niigata (+7.3%) rose in winter, along with small increases in Ehime, Tottori and Kumamoto (+0.4% to +2.8%) — a year ago, winter meant “off-season rates.” By January 2026 that had reversed, to Fukuoka +7.1% and Saga +18.3%. January-over-January YoY also shows double-digit growth across the board — Fukuoka +25.7%, Hokkaido +26.5%, Oita +19.9%, Okinawa +19.9%, Kagawa +22.9% — so it is reasonable to read Korea’s 1.176 million January visitors as reflected in host-prefecture rates.

Source: Compiled by HotelBank Editorial Team from MetroEngines Research

Source: Compiled by HotelBank Editorial Team from MetroEngines Research. ADR is each month’s estimated settled ADR (median of covered properties in the prefecture, tax-excluded equivalent); all values are for closed past months. N = number of properties in the estimate (January 2026)

Onsen, Ski, City — Three Types, Three Ways January Rates Rise

Splitting host prefectures further by property type, the way winter rates rise falls into three patterns. The first is ryokan in onsen prefectures. Oita ryokan rose from ¥18,100 to ¥21,500, +18.9% (N=304 properties); Saga ryokan from ¥19,300 to ¥24,000, +24.7% (N=88); and Kagoshima ryokan +31.6% (N=82) — all outpacing their prefecture-wide growth. Nagasaki ryokan jumped +65.7%, but with a small base of N=70 properties it should be read with a wide margin. January-over-January YoY was also strong — Oita ryokan +15.7%, Saga ryokan +15.5%, Kagoshima ryokan +23.7% — suggesting that capturing higher rates in winter is becoming established for onsen-prefecture ryokan.

The second is snow-season resorts. Hokkaido resort hotels rose from ¥14,100 to ¥19,100, +35.8% (N=128 properties), and Nagano resorts from ¥16,600 to ¥20,900, +25.8% (N=132) — the largest autumn→winter increases of the three types. However, Hokkaido resorts’ January YoY was only +1.3%, a gap from Nagano’s +22.9%. Hokkaido’s snow resorts already had high January rates a year ago, and in 2026 they maintained that level. The snow-season demand JNTO cited as a January growth driver shows up on the rate side as “maintained” in Hokkaido and “added on” in Nagano.

The third is city hotels, which absorb volume while keeping rates flat. Fukuoka business hotels went from ¥10,400 to ¥10,300, −1.1% (N=338 properties), and Hokkaido city hotels −0.2% (N=75) — essentially at autumn levels — with only Fukuoka city hotels rising, +7.2% (N=38). That said, January-over-January YoY grew by double digits in every case — Fukuoka business +14.4%, Fukuoka city +28.8%, Hokkaido city +29.4% — so the winter level itself has stepped up sharply in a year. The likely structure is that urban hotels already command high rates from autumn business-travel and event demand, so even with Korean guests added in winter, the seasonal gap does not exceed autumn. We also cover the difference between Fukuoka’s business and city hotels in Fukuoka Hotel ADR H1 2026: City Q1 +18.9% to Q2 +3.4%, Business +11.8%.

Source: Compiled by HotelBank Editorial Team from MetroEngines Research

Source: Compiled by HotelBank Editorial Team from MetroEngines Research. Segments follow each property’s registered category; ADR is estimated settled ADR (median, tax-excluded equivalent). N = number of properties in the estimate

Implications for Operators — How the Three Types Should Capture Winter Break and Seollal 2027

Given all this, it is practical to design winter sales for the Korean market along two axes: “when” and “by which type.” On “when,” 2026 confirmed that January peaks even without a holiday. January 2026’s 1.176 million came in a month without the Lunar New Year (Seollal), and exceeded the 1.086 million of February, which included the holiday. Korea’s Lunar New Year moved from January 28 (Tue)–30 (Thu) in 2025 to February 16 (Mon)–18 (Wed) in 2026, and in 2027 it will be the four days of February 6 (Sat)–9 (Tue). The 2026 pattern suggests capturing all of January through the build-up of winter break and weekends, and treating the early-February holiday as separate demand stacked on top. The 2027 calendar, in which four markets’ holidays overlap on February 6–9, and February statistics by prefecture are covered in Lunar New Year 2027: 4 Markets Converge Feb 6–9, Hokkaido ADR 1.18x.

On “by which type,” the room for growth differs across the three types. Onsen-prefecture ryokan have now shown January rate lifts for two years running, as YoY of +15.7% in Oita and +15.5% in Saga attests. In Kagoshima, Oita, Saga and Yamaguchi, where Korea’s share reaches 50–70%, one approach is to lock in January weekends and the days around holidays early as winter peak rates, and fill weekdays with value propositions such as multi-night stays and half-board (two meals). As shown in Winter Inbound Beyond Snow: Kagoshima, Oita, Saga and Yamaguchi, winter room occupancy at ryokan in these four non-snow prefectures stays around 40%, leaving upside in both rate and occupancy.

Among snow-season resorts, Hokkaido maintained its high January level while Nagano added +22.9%. Korean stays concentrate in Hokkaido at 316,120 guest nights, whereas Nagano is still thin at 4,110 (1.7% of the prefecture’s foreign guest nights), with Australia (55,550) and Taiwan the main players. For ski areas in Nagano and Niigata, Korea could become additional demand that lifts winter rate levels further. City hotels are the volume-absorbing type; with January YoY levels stepping up — Fukuoka business +14.4%, city +28.8% — there is room to rethink the very premise of pricing winter as off-season and to treat weekends and holidays on par with autumn event days. Read together with our observation that prefecture-level occupancy in Kyushu stayed near normal during the Chuseok holiday (Chuseok 2026: Did Japan’s Hotels Fill Up?), the Korean market appears more likely to translate into revenue when designed around “winter as a whole” rather than “the momentary spike of a holiday.”

Summary

Korean arrivals reached 1,176,000 in January 2026, the first time any market has topped 1.1 million in a single month, and growth continued through September at +21.6% cumulative. The annual peak is in winter centered on January, with January and February alone accounting for 27.4% of the January–September total. In the Japan Tourism Agency’s January 2026 data, Korea ranked first by nationality with 2,047,670 guest nights; by volume, 67.2% went to Tokyo, Osaka, Hokkaido and Fukuoka, while by share, western Japan’s onsen prefectures and Shikoku/San’in depend heavily on Korea — Kagoshima 68.9%, Ehime 65.9%, Yamaguchi 61.0%, Oita 48.0%, Fukuoka 46.1%. Estimated settled ADR in host prefectures was higher in winter than autumn from October 2025 to January 2026 — Oita +18.1%, Saga +18.3%, Kumamoto +13.6%, Fukuoka +7.1% — and given that the autumn→winter change was negative a year earlier, winter is shifting from off-season to high season. With onsen-prefecture ryokan and ski-area resorts capturing rate and city hotels capturing volume, the starting point for winter 2026–27 is a design that captures January as a whole and stacks the early-February Seollal holiday on top.

Frequently Asked Questions