›

›

Forest and Onsen

Forest and Hot Spring Together: The Japanese Habit of Walking Under Cypress and Then Bathing

Two old Japanese ways of resting the body meet in the Kiso valley. Cypress air, hot water, and what is known.

Kinosaki Onsen Open Air Bath

Many of Japan's prominent hot springs are often hidden in the valleys of these forests

On this page

No headings found on page

On this page

No headings found on page

Two old habits, one valley

Japan has two old ways of resting the body, and in the Kiso valley of Nagano they stand within a few kilometers of each other. One is the slow walk under cypress at Akasawa, where a forestry office proposed the word 森林浴 (shinrin-yoku) in 1982. The other is 湯治 (tōji), the hot-spring cure: a stay of days or weeks at an onsen, soaking several times a day in water that comes out of the ground already warm, with rest between.

What follows is about the two together, and about what is and is not known when they meet. It uses the documented Japanese word, 湯治, for the documented tradition, and it describes the pairing with the forest in plain English, because no Japanese source names that pairing and we will not be the first. The original contribution here is the synthesis, not a term.

The forest half is measured, modestly, and the science record reads the studies one at a time. The bath half is older, and most of its documentation is history; the physiology that exists is small, recent and read below with its sample sizes attached. What follows keeps the kinds of knowing in separate hands, which is the only honest way to hold them.

湯治: the hot-spring cure, as Japan practiced it

The National Diet Library’s own essay on the subject gives the shape of the tradition. Hot springs appear in the oldest Japanese chronicles: the 日本書紀 (Nihon Shoki) records imperial visits to springs in the seventh century; the surviving fragment of the 伊予国風土記 places Prince Shōtoku at a spring in 596; the 出雲国風土記 calls its waters 神湯, divine water. By the Heian period Sei Shōnagon was praising three springs in the 枕草子, and in the fifteenth century the Zen monk Zuikei Shūhō wrote a 温泉行記, a record of a journey to the baths.

Warlords used the water for their wounded; the essay names Takeda Shingen, who kept springs for his soldiers. By the Edo period the practice had reached ordinary people, with guidebooks such as the 七湯栞 for the seven springs of Hakone, and in 1738 the physician 香川修徳 (Kagawa Shūtoku) analyzed the springs medically in his 一本堂薬選. Tōji was, by then, a thing farmers did between harvests as well as a thing lords did.

The Edo custom set the length. One cycle, 一巡り (hitomeguri), was seven days and the minimum unit; the National Diet Library essay records that at least three cycles, 三巡り, twenty-one days, were understood to be needed for a proper cure. The bath was taken several times a day, in short soaks, with long rests and plain food between; the inns that served it, 湯治宿, were built for people who cooked for themselves and stayed for weeks.

The modern medical society keeps almost the same number. The Ministry of the Environment’s guide to safe onsen use, supervised by the 日本温泉気候物理医学会 (Japanese Society of Balneology, Climatology and Physical Medicine) and dated March 2019, says that two to three weeks is the appropriate length for a course of spa treatment if its full effect is wanted. It names the old hazard too: 湯あたり, the malaise, sleeplessness or stomach upset that can appear about three days to a week into a course. Those are the tradition’s own timings, written down by the people who still practice it, and they are stated here as guidance rather than as findings.

Waterfall dropping into a stream bed stained deep orange by mineral deposits, surrounded by green forest at Kusatsu, Gunma

What counts as an onsen today is a matter of law. The 温泉法 (Onsen Law) defines one as water, mineral water, steam or other gas issuing from the ground at 25 °C (77 °F) or above at the source, or containing at least one of a schedule of substances above set concentrations, such as 1,000 mg/kg of dissolved solids or 250 mg/kg of free carbon dioxide. Twenty-five degrees is cool to the hand; the water in the baths of these towns is heated by the earth to well above it, and by the inn to a comfortable soak.

The Ministry of the Environment keeps the tradition on the national books. It designates 国民保養温泉地, national recuperation spa areas, and since 2018 has run a 新・湯治, “new tōji,” program that reframes the stay as time spent in the whole environment of a spa town, not only in the water. The ministry says it gathered nearly 20,000 survey responses on that program between fiscal 2018 and 2023; what those surveys found is in a summary report we have not read. That is the state’s own version of the synthesis described here: the bath, the walk, the air, the town.

What is known about heat and the body

This section is written under the property’s rule: a physiological claim carries a named study, a sample size, a design and one stated limitation, or it does not appear. For hot-spring bathing the studies that pass are few, small and mostly Japanese, and the larger numbers come from cohorts that can show association and nothing more. Here they are, one at a time, with what each cannot say beside what it can.

The stress-hormone review

The review most often cited for the bath’s effect on stress is Antonelli and Donelli’s 2018 systematic review in the International Journal of Biometeorology, which searched PubMed, Embase and the Cochrane Library for studies measuring cortisol in people taking balneotherapy or spa therapy. It found 15 studies with 684 participants in all. Of five studies of bathing alone, two reported a significant change in cortisol in healthy people and three reported none in patients with rheumatic disease; of ten studies of spa therapy more broadly, eight reported significant variation. The authors’ own conclusion is that bathing “may have the potential to influence cortisol levels in healthy subjects,” and their stated limitation is that “further investigation is needed because of limited available data.”

Pale turquoise hot spring pool with steam rising off it among autumn scrub at Kazuno, Akita

Water coming out of the mountain at close to boiling, and cooling in about fifty meters.

Two things about that review matter for a reader. It pooled no numbers, so there is no single figure for how much cortisol moves, and we will not invent one. Its full text sits behind a subscription we did not buy, so what is reported here is the abstract, which is where the authors put their counts and their caution. The same authors published an updated review in 2024 that we have not read. Fifteen studies of 684 people, most of them patients, is the whole of the cortisol literature on the bath, and it is smaller than the forest’s.

The Japanese bathing experiments

The Japanese Society of Balneology, Climatology and Physical Medicine publishes a journal, and its bathing experiments are the closest thing to the forest field studies: small, controlled, short, and honest about it. Four are read here from their English abstracts; none of the abstracts prints an effect size, so none is printed below.

Mori and colleagues, 2016, had 16 people bathe for fifteen minutes at 38 to 39 °C (100 to 102 °F) in two tubs on separate occasions, one of hinoki cypress and one an ordinary modular bath, with a Holter monitor running throughout and saliva taken before and after. Salivary cortisol fell after bathing in both tubs, and the high-frequency heart-rate variability that marks parasympathetic activity was higher after both. The hinoki tub scored higher for pleasantness and lower for felt fatigue on the questionnaires. The design is within-subject with no unbathed control, and the abstract states no limitation; what it shows is that a warm bath of any kind moved the same two markers the forest studies use, and that the cypress changed how the bath felt rather than what the saliva did.

Kubo and colleagues, 2014, had 18 habitual shower-users bathe at 41 °C (106 °F) for ten minutes on fourteen consecutive nights in a crossover with a two-week washout, and repeated the design in a second group of 20. The outcome was a mood questionnaire, on which depression, fatigue and confusion scores moved toward the population mean after the bathing fortnight. The measures were subjective only.

Shimasaki and colleagues, 2018, had 10 healthy men aged around 25 bathe for ten minutes at 42 °C (108 °F) in a large hot-spring plunge bath and in a home tub, and found deep body temperature rising significantly from the third minute in the spring. The two tubs differed in volume as well as water, so the study cannot separate the spring from the size of the bath.

Rough stone open-air hot spring pool set among boulders and golden birch forest at Myoko, Niigata

An unbuilt bath on the mountain, which is the oldest form of all of this.

Miwa and colleagues, 2021, is the one to read before getting in. Ten young adults bathed for twenty minutes at 40 °C (104 °F) in summer and 42 °C (108 °F) in winter, and systolic blood pressure rose significantly immediately after bathing in both seasons, with the cardiovascular load greater in the hotter winter bath. Temperature and season are confounded by the design, and ten people is ten people. It is nonetheless the finding that underwrites the medical society’s forty-two-degree ceiling, below, and it is why we put the bath after the walk and not before it.

One further Japanese trial is better designed and measures something else. Goto and colleagues, 2018, randomized 38 people in a crossover to two weeks of a ten-minute tub bath at 40 °C against two weeks of showers, and found better scores for fatigue, stress and mood after the bathing fortnight. The authors’ own limitation is that every measure was subjective and no physiological marker was taken. That is a fair description of most of what exists on the bath: people say they feel better, and the small physiological studies point the same way without yet putting a number on it.

The big numbers, and what they cannot say

The largest figures about Japanese bathing come from cohorts, and a cohort can show that people who bathe a certain way are, on average, different from people who do not. It cannot show that the bath made them so.

Ukai and colleagues, 2020, in Heart, followed 30,076 adults in the Japan Public Health Center cohort from 1990 to 2009 and found that those who took a tub bath almost daily had a hazard ratio for cardiovascular disease of 0.72 (95% confidence interval 0.62 to 0.84) against those who bathed twice a week or less, with similar figures for coronary disease and stroke. The authors list their limitations plainly: bathing frequency was asked once, at baseline; frequent bathing “may be a behavioral marker for higher socioeconomic conditions”; and residual confounding cannot be excluded.

Kohara and colleagues, 2018, in Scientific Reports, measured 873 middle-aged and older people in Ehime and found that those who bathed in hot water five or more times a week had lower arterial stiffness, central pulse pressure and B-type natriuretic peptide. The study is cross-sectional, which the authors say “makes it impossible to infer any causal relationships,” and it did not measure water temperature, which was classed by the bather’s own word.

Waterfall dropping over a wall of columnar basalt into a dark pool framed by autumn trees at Kazuno, Akita

The fall comes off a basalt wall, which is the volcano showing its working.

The famous Finnish sauna cohort is often cited beside these, and it does not belong here. Laukkanen and colleagues, 2015, followed 2,315 men for a median of 20.7 years and found lower cardiovascular and all-cause mortality among frequent sauna users, and then wrote that the results apply to a dry Finnish sauna at 80 to 100 °C and cannot be “directly applied to steam rooms, hot tubs.” We take them at their word. Nor does the laboratory heat-therapy literature transfer: the best-known trial, Brunt and colleagues in 2016, immersed 20 sedentary young adults at 40.5 °C for an hour four to five times a week for eight weeks, keeping core temperature above 38.5 °C, and that is a protocol, not an onsen afternoon.

What the cohorts leave a reader with is this. Japanese people who bathe in a tub most days have, on average, somewhat lower rates of heart disease and stroke than those who rarely do, and nobody knows how much of that is the bath. It is an association worth having and not a claim we will make into advice.

The medical society’s rules, which are advice

What can be stated without hedging is the practical guidance the tradition’s own medical society gives, because it is advice rather than a finding. The Ministry of the Environment’s 2019 guide, supervised by the society, says that older people and people with high blood pressure, heart disease or a history of stroke should avoid baths of 42 °C (108 °F) and above; that a beginner should soak for three to ten minutes, once or twice, and a habitual bather for fifteen to twenty minutes, two or three times; that water should be poured over the hands and feet and then the whole body before entering, and the body washed; that fluids should be taken before and after; and that bathing should be avoided just before or after a meal, after drinking, when very tired, and for thirty minutes after exercise.

The guide gives no single recommended temperature, and we do not supply one. It does say that people with weakened heart or lung function should prefer a half-body or partial bath, that the elderly, children and people with disabilities should not bathe alone, and that older people with sensitive skin should stay out of acidic and sulfur springs altogether. Those are the rules of a craft, written down by the people who practice it.

Wide multi-tiered waterfall spreading across a dark rock face into a shallow pool at Nichinan, Miyazaki

Goju-no-taki, the fifth and last fall, two and a half hours up the gorge.

They also settle the order of the day we turn to below. A thirty-minute gap after exercise is the guide’s own instruction, and a forest bathing walk is, by the 2025 Tokyo trial’s accelerometers, about 2.2 METs of effort, lighter than ordinary walking but effort all the same. Miwa’s ten subjects, above, are the reason the gap is there.

The one thing the forest studies say about the bath

The Japanese immune studies deliberately kept the water out. Li’s 2010 review of the three-day forest trips states that subjects did not take hot-spring baths or eat special foods during the trips, so that the forest could be measured on its own. That is why nothing in the forest literature can be read as a result about the bath, and why the section on what has not been measured, at the end of this page, is as long as it is.

What is known about cypress air

Forest air over Japanese conifers carries measurable α-pinene and β-pinene, nearly absent from city air; Qing Li’s group sampled it during the trips reported in 2008 and in the 2009 hotel-room study. That is chemistry and it holds. The step from the presence of terpenes to a cause of anything in the body is the step the science record walks slowly, with the twelve-man hotel-room study read plainly, and we do not repeat it.

One fact from that literature belongs here because of what it says about the bath. Hinoki, the cypress of Akasawa, is also the wood of the traditional Japanese bathtub, and the smell of a hinoki bath before the water goes in is the smell of the forest approach road. The 2009 study vaporized hinoki stem oil in a Tokyo hotel room for twelve men over three nights and compared them with themselves; it did not test a wooden bath, hot water, or the two together.

The nearest thing to that test is Mori’s 2016 hinoki-tub study above, and it is worth being exact about what it found. Sixteen people, warm water, two tubs: cortisol and heart-rate variability moved the same way in both, and only the questionnaires told the cypress from the plastic. On that evidence the hinoki tub is a better place to sit, not a different physiology. Whether the wood’s terpenes reach the water in any measurable amount has not been tested in any study we have found, searching to September 2026.

Small waterfall falling into a clear green pool between mossy boulders at Sasaguri, Fukuoka

The fall is three meters and the pool below it is deeper than it looks.

Where the two are already paired

Every certified forest in this registry has a hot-spring town within reach, and that is not a coincidence of the selection; it is the geography of Japanese mountain towns. The site records carry each pairing with its access, and this section only names them.

Akasawa’s town, Agematsu, is one local stop, eight or nine minutes, from 木曽福島 (Kiso-Fukushima), the old post town whose baths look at the river; the same Chūō-line station is where the limited express from Nagoya stops, eighty-eight minutes out. Shiga Kōgen’s five certified roads sit on the plateau above 渋温泉 (Shibu) and 湯田中 (Yudanaka), in the same town of Yamanouchi and on the same bus line, with a string of mountain springs on the plateau itself where the inns take day bathers. Kusatsu’s three roads are at the edge of the onsen town whose central square runs with hot water; Kirishima’s four roads lie between two onsen districts the Society’s own page names, and one road passes what that page calls the oldest rock bath in Kirishima.

The two uncertified forests pair with the strangest baths. Yakushima’s cedars sit above 平内海中温泉 (Hirauchi Kaichū Onsen), rock pools in the sea that the town’s page says appear for about two hours either side of the day’s two low tides, for a contribution of ¥300 (about $2), mixed bathing, no swimsuits, a towel or bathing robe allowed. The Kumano Kodō’s Nakahechi route ends at 那智勝浦 (Nachi-Katsuura), where the baths include one in a sea cave, and its middle passes the springs at Yunomine and Kawayu that pilgrims used for centuries.

Two forests have no onsen-town pairing documented in our sources: Shinano-machi and Iiyama, where the bath is the inn’s. We do not invent one for them.

A day that does both, in order

The tradition put the walk first and the water after, and the medical society’s guidance gives the reason in one line: no bath for thirty minutes after exercise, and none just before or after a meal. A forest bathing walk is light effort, but it is effort, and Miwa’s ten bathers with their blood pressure up on leaving the water are the reason the rule stands.

Waterfall fanning out over a dark rock face into a shallow pool in woodland at Nanasawa, Kanagawa

The fall at the head of the valley, running hard after rain.

So the morning is the forest. The measured session in the Japanese field experiments was about fifteen minutes of sitting and looking and about fifteen minutes of walking, on a wide path; the protocol record sets that out step by step, with a longer sit than most visitors allow themselves. The Forest Therapy Society’s own guidance is more generous, a half day of about two hours and 2.5 km (1.6 mi) or a full day of about four hours and 5 km (3.1 mi), and it says the time should not be watched. Go early, before the parking lot fills and before the day warms, and take the resin smell as it comes on the approach road.

Then the lunchbox, if the town makes one, eaten slowly in the wood; then the bus or the walk down to the town; then at least half an hour of nothing before the bath. The medical society’s guide asks for water before and after, and for the pour over hands and feet and body before entering. A beginner soaks three to ten minutes, once or twice; someone used to it, fifteen to twenty, two or three times, with rests on the edge between, and never above 42 °C (108 °F) for the old, the hypertensive or anyone with a heart condition.

That rhythm, a little, slowly, and again, is the same rhythm the forest asks for. It is also the Edo cycle in miniature: the seven-day 一巡り compressed to an afternoon. We will not claim the afternoon does what the three weeks did, because nobody has measured either, and the medical society itself says the full effect wants two to three weeks.

The evening is the inn’s. The Kiso valley, the Shibu bath villages and the Kusatsu square have been doing this for as long as there have been both a forest and a spring, and the day described here is theirs, written down.

A narrow waterfall dropping down a mossy rock face into a shallow pool.

Water comes off these slopes year round.

What no one has measured

No published study we have found has tested a forest walk and a hot-spring bath together and reported what the combination does to the body. The search was run on 29 September 2026 in Japanese and English, across the balneology society’s journal and the general literature.

What the searches return instead is instructive: forest-only trials, including Kondo and colleagues’ 2011 study in the balneology society’s own journal of 48 healthy people in their mid-sixties, half of them women, who spent two hours in a forest against two hours resting indoors and showed lower adrenaline and cortisol and no change in natural killer activity; a Beppu spa-health research grant with no forest arm; and a 2021 spa-stay program of yoga, walking and bathing in 31 women, single-arm, with no way to tell the bath from the rest.

The forest studies excluded the bath by design; the balneology literature, so far as we have read it, has no forest arm. The Forest Therapy Society’s own how-to page recommends a two-night, three-day stay for the immune measures, which is the length of the Nippon Medical School trips, and says nothing about the water. The synthesis described here is a tradition, not a result.

That is worth stating as plainly as the property states anything, because the pairing is the kind of thing that gets a coined name and a mechanism within a year of being written up. There is no mechanism here. There is a valley with a cypress forest at the top and a bath at the bottom, and a very old habit of doing them in that order.

What would have to exist for this section to change is the same as for the forest alone: people measured before and after, against a control that had one and not the other, over more than an afternoon, in more than twelve of them. Until then, the bath after the walk is offered here as what it is: the best part of the day, and not a treatment for anything.

Verified by The Japan Onsen Society

  • 国立国会図書館「本の万華鏡 — 湯治場としての温泉」 (National Diet Library, on tōji; the 一巡り / 三巡り cycles and the named classical sources) — ndl.go.jp

  • 環境省「温泉の定義」 (Ministry of the Environment, the Onsen Law definition) — env.go.jp

  • 環境省『あんしん・あんぜんな温泉利用のいろは』監修 日本温泉気候物理医学会, March 2019 (read in full 29 September 2026) — env.go.jp

  • 環境省「新・湯治による効果等について」 (the New Tōji program) — env.go.jp

  • 環境省「国民保養温泉地」 (national recuperation spa areas) — env.go.jp

  • Antonelli M, Donelli D. Int J Biometeorol. 2018;62(6):913–924 (abstract only; paywalled. 15 studies, 684 participants) — doi.org

  • Antonelli M, Donelli D, et al. Int J Biometeorol. 2024;68(10):1909–1922 (identified, not read) — doi.org

  • 森 康則ほか, 日本温泉気候物理医学会雑誌 2016;80(2):66–72 (n=16, hinoki vs modular tub) — doi.org

  • 久保 達彦ほか, 日本温泉気候物理医学会雑誌 2014;77(5):511–512 (n=18 and 20, crossover, 41 °C) — doi.org

  • 島崎 博ほか, 日本温泉気候物理医学会雑誌 2018;81(2):63–69 (n=10 men, 42 °C) — doi.org

  • 三輪 ちほかほか, 日本温泉気候物理医学会雑誌 2021;84(2):65–74 (n=10, blood pressure on exit) — doi.org

  • 西村・山仲, 日本温泉気候物理医学会雑誌 2021;84(2):75–80 (n=31 women, single-arm spa program) — doi.org

  • 近藤照彦ほか, 日本温泉気候物理医学会雑誌 2011;74(3):169–177 (n=48, forest vs indoor rest, no bathing arm) — doi.org

  • Goto Y, Hayasaka S, Kurihara S, Nakamura Y. Evid Based Complement Alternat Med. 2018;2018:9521086 (n=38 crossover; subjective measures only) — doi.org

  • Ukai T, Iso H, Yamagishi K, et al. Heart. 2020;106(10):732–737 (JPHC cohort, n=30,076; full text read) — heart.bmj.com

  • Kohara K, et al. Sci Rep. 2018;8:8687 (n=873, cross-sectional; full text read) — nature.com

  • Laukkanen T, et al. JAMA Intern Med. 2015;175(4):542–548 (n=2,315; the authors state the results do not transfer to hot tubs) — doi.org

  • Brunt VE, et al. J Physiol. 2016;594(18):5329–5342 (abstract only) — doi.org

  • 森林セラピーソサエティ「森林セラピーの楽しみ方」 (half day 2 h / 2.5 km; full day 4 h / 5 km) — fo-society.jp

  • Li Q. Environ Health Prev Med. 2010;15:9–17 (subjects took no hot-spring baths during the forest trips) — link.springer.com

  • Li Q, Morimoto K, Kobayashi M, et al. Int J Immunopathol Pharmacol. 2008;21(1):117–127 (forest and city air sampling) — journals.sagepub.com

  • Li Q, Kobayashi M, Wakayama Y, et al. Int J Immunopathol Pharmacol. 2009;22(4):951–959 (the hinoki-oil hotel room) — journals.sagepub.com

  • Ochiai H, Inoue S, Masuda G, et al. Sci Rep. 2025;15:3272 (the 2.2 METs figure) — nature.com

  • 屋久島町「平内海中温泉」 (updated 10 March 2026) — town.yakushima.kagoshima.jp

Yen figures are converted at roughly ¥150 to the US dollar. Verified 18–29 September 2026.

No headings found on page