In 1982, a Japanese government forestry body launched a national health program with an unusual target audience: office workers. The proposal was that walking slowly in a forest would help them handle stress. It came with a name, shinrin-yoku, which translates roughly as forest bathing.
The name traveled. Four decades later you can book a guided forest bathing session in most US states, often for more than the price of a yoga class.
The research came afterward. By the account of Qing Li, the Nippon Medical School researcher whose studies made the practice internationally famous, the early 1980s program was “based only on common sense and the intuitive idea that being in the beautiful green forests of Japan would be good for us.”
That order matters, and it is not an insult to the practice. Plenty of useful things begin as intuition. But it means the studies arrived to evaluate a program that already existed, in a country where it was already popular, often run by people who already believed in it. That shapes what the evidence can carry, and it is the reason this article spends more time on study design than on benefits.
A forestry agency, a stressed workforce, and a phrase
Start with the geography. Forests cover about 67 percent of Japan’s land area, so a national program built around forest visits was working with an asset the country already had in abundance.
Two published accounts credit the origin slightly differently, and the difference is worth stating rather than smoothing over. A 2010 paper in Environmental Health and Preventive Medicine by Bum Jin Park and colleagues at Chiba University says the term shinrin-yoku was coined by the Japanese Ministry of Agriculture, Forestry and Fisheries in 1982. Li’s 2022 review in the same journal says a national health program for shinrin-yoku was introduced that year by the Forest Agency of Japan, aimed at stress management for workers.
The occupational context was explicit. Li’s review cites figures from Japan’s Ministry of Health, Labour and Welfare showing that the share of workers reporting anxiety and stress ran above 50 percent in 1982 and was still around 60 percent three decades later.
It worked as a public campaign. A 2003 opinion poll cited in the same review found that 25.6 percent of Japanese respondents had been on a shinrin-yoku trip.
The research infrastructure arrived two decades after the name. The Japanese Society of Forest Therapy was founded in 2004. The Ministry of Agriculture, Forestry and Fisheries ran a project from 2004 to 2006 to investigate whether forests had measurable therapeutic effects. Li’s team conducted its first shinrin-yoku study in 2005, at Iiyama in Nagano prefecture.
So the sequence is: forestry policy, then popularity, then science. That is the opposite of how a medical intervention usually reaches the public, and it explains a lot about the literature’s shape.
What a session involves
Li’s description is the clearest one in the published literature, and it is notable for what it rules out. This is not exercise, hiking or jogging. It is being in a forest and taking it in through sight, smell, hearing, touch and taste: leaf color and forest views, the smell of trees, birds and wind in the branches, the feel of bark, the taste of forest air and forest food.
In practice, a guided session looks like a slow walk covering very little ground, broken by long stops, with a guide offering prompts to notice specific things, usually some time sitting still, and sometimes tea at the end. Pace is the point. Distance is not.
There is no standardized protocol, and that is a finding rather than an oversight. A 2026 review in the journal Medical Sciences by Arnab Bandyopadhyay and colleagues at the University of Naples Federico II concluded that standardized intervention protocols are still needed, and that heterogeneity in design, exposure and outcome measures limits how well studies can be compared.
The exposures in the research are also shorter than most commercial sessions. Park’s field experiments used about 16 minutes of walking followed by about 14 minutes of seated viewing. Li’s trips covered roughly 2.5 kilometers a day, a distance chosen because it resembles an average working day’s walking, spread across two-hour walks on a three-day trip.
The natural killer cell studies, in detail
Four studies from a single laboratory produced the immune claims that travel furthest. They deserve to be read closely rather than cited as a block.
Natural killer cells are a type of white blood cell that can destroy certain infected or abnormal cells without needing prior exposure to them. What these papers call NK activity is a laboratory measurement: how effectively cells drawn from a blood sample kill target cells in a dish.
2007, 12 men. Published in the International Journal of Immunopathology and Pharmacology, this study recruited 12 healthy men aged 37 to 55 from three large Tokyo companies and sent them on a three-day, two-night trip to forests in Nagano. Eleven of the 12 had higher NK activity afterward, roughly a 50 percent increase. There was no control group. Each man was compared with his own reading taken on an ordinary working day.
2008, another 12 men, plus a city trip. The follow-up used 12 healthy men aged 35 to 56 and added a matched city trip and longer tracking. NK activity stayed elevated for more than seven days, and Li’s 2022 review reports some measures still raised at 30 days. The city trip produced no increase. Two phytoncides, alpha-pinene and beta-pinene, were detected in forest air and almost none in city air.
2008, 13 nurses. Published in the Journal of Biological Regulators and Homeostatic Agents, this study repeated the design with 13 healthy female nurses aged 25 to 43. Results pointed the same way, with urinary adrenaline and noradrenaline also lower.
Now the detail that most coverage leaves out. What was the city walk? According to Li’s own 2022 account, the city group walked for two hours around the Nagoya Baseball Dome in the morning and two hours around Nagoya Airport in the afternoon, in areas with almost no trees.
That is a defensible way to hold walking constant while removing trees. It is not a comparison between a forest and an ordinary city park, and it cannot tell you how a forest stacks up against the quiet green park six blocks from your house.
Add the participants up. The four best-known immune studies in this field came from one research group at Nippon Medical School and together enrolled fewer than 50 people. All were healthy. All were Japanese. None were randomized, none were blinded, and most were compared against their own earlier readings rather than a control group.
That is not bad science. It is a pilot literature, and pilot work is supposed to look like this. The problem arrives when pilot results get quoted as settled.
Phytoncides, and the hotel room experiment
Phytoncides are volatile compounds that trees release into the air. Alpha-pinene and beta-pinene, the two most often measured in this research, are terpenes, and they are a large part of what a conifer forest smells like.
The cleverest study in the set tried to separate the smell from the forest. In 2009, Li’s team had 12 healthy men aged 37 to 60 stay three nights in an urban hotel, from 7 p.m. to 8 a.m. A humidifier vaporized stem oil from hinoki cypress in the room overnight, and alpha-pinene and beta-pinene were measured in the room air. NK activity and the proportion of NK cells rose, and urinary adrenaline and noradrenaline fell.
Isolating one variable that way is good thinking. It also inherits every limitation of its siblings. There was no sham room, participants could obviously smell the oil, there were 12 of them, and the comparison was once again their own values from a normal working period. A study nobody can blind, in a group that small, measuring a laboratory marker, is a reason to run a bigger trial rather than a reason to conclude something.
Li’s own review contains a useful example of how unsettled the biomarker work still is. His team found that forest trips raised serum levels of one hormone marker, DHEA-S, while a Korean forest program reported the opposite direction, and Li writes that the effect needs further research. When two laboratories get opposite signs on the same marker, the marker is not established.
Why a mechanism study cannot stand in for a health outcome
The phrase anti-cancer proteins appears in the titles of several of these papers, and it is the single most misread thing in the subject.
The proteins involved, perforin, granulysin and granzymes A and B, are the molecular tools immune cells use to kill target cells. Finding more of them expressed in a blood sample after a trip is a real measurement about blood.
It is not a measurement about disease. No trial has randomly assigned people to forest visits and then followed them for any disease outcome. The gap between a cell-level marker and a clinical result is where most health claims in any field go wrong, and there is no shortcut across it.
The most reliable evidence for that gap comes from the field’s own leading researcher. Among the goals Li sets out for future international work in his 2022 review are verifying preventive effects on non-communicable diseases around the world and establishing forest bathing as a treatment for some of them. Those are listed as work still to be done, by the person best placed to know.
The honest tier for the immune findings, then, is preliminary research. Not disproven. Not established. Interesting enough to fund a proper trial, not strong enough to carry a health claim.
What is reasonably supported: short-term stress measures
The strongest part of this literature is the least dramatic part.
Park’s 2010 paper remains the largest of the classic physiological studies: field experiments in 24 forests across Japan, 12 participants at each site for 280 in total, mean age about 22. Each group spent one day in a forest area and the next in a city area, so every participant acted as their own control on consecutive days. Forest sessions were associated with lower salivary cortisol, lower pulse rate, lower blood pressure, greater parasympathetic nervous activity and less sympathetic activity.
A 2019 systematic review and meta-analysis by Michele Antonelli and colleagues in Parma, Italy, screened 971 articles, included 22 in the review and pooled 8. Salivary cortisol was significantly lower in forest groups than urban groups, both before and after the sessions. Their own framing is the interesting part: writing in the International Journal of Biometeorology, the authors conclude that “anticipated placebo effects can play an important role in it.” They also say the available data are limited and more research is advised.
The same group published an umbrella review in 2021 in the International Journal of Environmental Health Research, covering 16 systematic reviews. Its conclusion is carefully graded: the best available evidence supports forest bathing as a complementary practice for promoting psychophysical well-being, while evidence for its use as a treatment for physical disease needs to grow before any clinical indications can be written.
A 2019 systematic review in Environmental Health and Preventive Medicine by Ye Wen and colleagues at Jiangxi Agricultural University in China took the measure of study quality directly. From 210 papers published between 2015 and 2019, 28 met inclusion criteria, and the authors scored them with the Downs and Black checklist. Randomized trials scored significantly higher than non-randomized ones, and their recommendation was to raise methodological quality and reduce risk of bias.
Put together: short-term reductions in self-reported stress and in several physiological stress markers are the best-supported finding here. Anything durable is not.
The control group is usually a city walk, and that is the core problem
In most of these studies the comparison condition is a walk in a city. That holds the walking constant, which is exactly what you want. It also changes several other things at once: noise, traffic, air pollution, crowding, visual complexity, novelty, and what the participant expects to feel.
The 2017 state-of-the-art review by Margaret Hansen and colleagues at the University of San Francisco, which screened 127 papers and included 64, names this directly. Studies generally cannot separate effects of the surroundings from effects of the participant’s activity level in each setting, and most say little about which of the five senses did any work.
That review is also the clearest published statement of the field’s demographic problem. Original study samples were often under 20 participants, drawn mainly from healthy young male university students, which makes generalizing difficult. Gender and age were frequently homogeneous. Research was concentrated in Japan and China. And because nearly every published study reports a positive result, publication bias has to be assumed rather than ruled out.
Expectation is not a footnote in this subject. Someone who traveled to Nagano for a forest study knows which group they are in. Antonelli’s review treats anticipated placebo as part of the measured effect rather than as noise to be removed, which is a more honest approach than pretending it can be subtracted.
What would move the evidence forward is straightforward to describe and hard to run: a forest compared with a quiet urban park matched on noise and air quality, with outcomes that are not self-reported, measured weeks later rather than the same afternoon.
You do not need a forest, a guide or a protocol
This is the finding most likely to annoy people on both sides of the subject, so it is worth stating precisely.
Li’s own review makes the point from inside the research group that built the field. Walking in Tokyo city parks lowered negative mood scores and showed a relaxing effect in both men and women. Those were city parks, not forests.
The dose question points the same way. Article 728 handles the population survey work on how much weekly nature contact tracks with good health, and why no agency has turned that into a recommendation. Within the forest literature itself, the exposures were modest. Park’s sessions put participants among the trees for well under an hour. Li’s trips were built around an ordinary working day’s walking distance, not an expedition.
The practical reading is that the ingredient with the most support, slowing down and paying attention outdoors for a while, does not require old-growth timber, a certificate or a fee. If a guided session is what gets you to do it, that is a genuine benefit of guided sessions, and worth paying for on those grounds. If the cost or the booking is a barrier, the evidence does not demand it. Article 732 covers the theories behind why short nature exposures do anything at all, and article 727 covers getting outside generally.
None of that is a dismissal. The thing Japan’s forestry program was built to deliver, a socially acceptable reason to stop and be outside for two hours, turns out to be cheap.
Five questions to ask before booking a guided session
- How long is the walk, how far does it go, and what is the terrain and grade?
- What happens if it rains, and what is the cancellation or reschedule policy?
- Where is the nearest bathroom, and how far is it from the start?
- Who is the guide, and what specifically does their certification require? There is no government license for this in the United States, so certification means whatever the issuing body decided it means.
- What else is bundled in, such as tea, a meal, bodywork or products, and what is that portion costing you?
Where a clinician belongs in this
The National Center for Complementary and Integrative Health draws a distinction that fits here exactly. Complementary care runs alongside the treatment a doctor prescribes. Alternative care takes that treatment’s place. Nothing in the published forest bathing research supports the second category, and the reviews cited above say so themselves.
Some specifics that matter more than a general disclaimer. Chest pressure, unusual breathlessness or dizziness on a slow walk is a reason for same-day medical attention, not a reason to try a different trail. A cough or wheeze that reliably shows up outdoors deserves a clinician’s assessment rather than a change of forest. Low mood, anxiety or sleep trouble that lasts more than two weeks, or that is getting worse, needs an appointment; a weekend in the woods is not an evaluation. If you are in crisis, call or text 988. And nobody should change a prescription because of anything in this literature.
The short version worth keeping
Strip the subject to what the studies support and you get one sentence: spending a couple of hours a week outdoors, slowly and with your attention on it, is associated with feeling better in the short term, and the forest itself is the least tested ingredient in that sentence.
Everything stacked above it, the natural killer cell results, the phytoncide work, the disease-prevention framing, is early research whose own authors describe it as needing larger and better-designed trials. You can keep the practice and still read the headlines accurately. Those two things were never in conflict.
Sources
- Park BJ, Tsunetsugu Y, Kasetani T, Kagawa T, Miyazaki Y.. “The physiological effects of Shinrin-yoku (taking in the forest atmosphere or forest bathing): evidence from field experiments in 24 forests across Japan. Environmental Health and Preventive Medicine, 2010. PubMed. DOI: 10.1007/s12199-009-0086-9.” doi.org
- Li Q.. “Effects of forest environment (Shinrin-yoku/Forest bathing) on health promotion and disease prevention: the Establishment of Forest Medicine. Environmental Health and Preventive Medicine, 2022. PubMed. DOI: 10.1265/ehpm.22-00160.” doi.org
- Li Q, Morimoto K, Nakadai A, et al.. “Forest bathing enhances human natural killer activity and expression of anti-cancer proteins. International Journal of Immunopathology and Pharmacology, 2007. PubMed. DOI: 10.1177/03946320070200S202.” doi.org
- Li Q, Morimoto K, Kobayashi M, et al.. “Visiting a forest, but not a city, increases human natural killer activity and expression of anti-cancer proteins. International Journal of Immunopathology and Pharmacology, 2008. PubMed. DOI: 10.1177/039463200802100113.” doi.org
- Li Q, Morimoto K, Kobayashi M, et al.. “A forest bathing trip increases human natural killer activity and expression of anti-cancer proteins in female subjects. Journal of Biological Regulators and Homeostatic Agents, 2008. PubMed. PMID: 18394317 (no DOI assigned).” pubmed.ncbi.nlm.nih.gov
- Li Q, Kobayashi M, Wakayama Y, et al.. “Effect of phytoncide from trees on human natural killer cell function. International Journal of Immunopathology and Pharmacology, 2009. PubMed. DOI: 10.1177/039463200902200410.” doi.org
- Antonelli M, Barbieri G, Donelli D.. “Effects of forest bathing (shinrin-yoku) on levels of cortisol as a stress biomarker: a systematic review and meta-analysis. International Journal of Biometeorology, 2019. PubMed. DOI: 10.1007/s00484-019-01717-x.” doi.org
- Antonelli M, Donelli D, Carlone L, Maggini V, Firenzuoli F, Bedeschi E.. “Effects of forest bathing (shinrin-yoku) on individual well-being: an umbrella review. International Journal of Environmental Health Research, 2021. PubMed. DOI: 10.1080/09603123.2021.1919293.” doi.org
- Hansen MM, Jones R, Tocchini K.. “Shinrin-Yoku (Forest Bathing) and Nature Therapy: A State-of-the-Art Review. International Journal of Environmental Research and Public Health, 2017. PubMed. DOI: 10.3390/ijerph14080851.” doi.org
- Wen Y, Yan Q, Pan Y, Gu X, Liu Y.. “Medical empirical research on forest bathing (Shinrin-yoku): a systematic review. Environmental Health and Preventive Medicine, 2019. PubMed. DOI: 10.1186/s12199-019-0822-8.” doi.org
- Bandyopadhyay A, Shah S, Roviello GN.. “Forest Bathing and Preventive Medicine: Immune Modulation, Stress Regulation, Neurocognitive Resilience, and Neurological Health. Medical Sciences, 2026. PubMed. DOI: 10.3390/medsci14010095.” doi.org
- National Center for Complementary and Integrative Health, National Institutes of Health. “Complementary, Alternative, or Integrative Health: What's In a Name?.” nccih.nih.gov