The Advice Everyone Already Believes
Step outside, feel better. That advice is so familiar it barely counts as advice, and its vagueness is part of its charm, because any walk counts, any park counts, any amount counts, and if a little is good, a lot must be better.
A team at the University of Exeter checked the dose, and the answer did not look like a slope. It looked like a cliff. People reporting less than two hours of recreational nature contact in the previous week looking, in the main adjusted analysis, no different from people reporting none at all, while people crossing two hours looked better and people going far beyond two hours did not keep getting better.
A Survey, Not a Prescription
The evidence comes from the Monitor of Engagement with the Natural Environment survey, a UK government survey run across England through the year with in-home, face-to-face interviews. The researchers used the two waves in which the health and wellbeing questions were asked consistently, 2014/15 and 2015/16, with the 19,806-person sample weighted to stand in for the adult population of England.
Nature contact meant recreational visits in the last seven days to parks, canals, nature areas, coasts, beaches, farmland, woodland, hills, and rivers. Two exclusions matter: routine shopping trips did not count, and neither did time in your own garden. The survey counted recreation, not views from a window, and not time outdoors at work.
Weekly minutes were not measured with a stopwatch. Interviewers recorded the visit count, then asked in detail about one visit chosen at random. Weekly contact was that visit length multiplied by the visit count, capped at seven visits. Health was a single question, "How is your health in general?", split into good or very good versus fair, bad, or very bad, and wellbeing was a single life-satisfaction question scored 0 to 10, split into high (8 to 10) versus low (0 to 7). Both measures are crude, and both are widely used because self-rated health tracks medical service use and mortality in other research. The study measured how people rated their lives, not their blood pressure or medical records.
The Cliff at Two Hours
Before adjustment, every category from 60 minutes upward looked better than zero. After adjustment for neighbourhood greenspace, deprivation, air pollution, demographics, employment, relationship status, physical activity, dog ownership, and survey year, the picture sharpened. The 1 to 59 and 60 to 119 minute categories lost significance, while significance appeared at 120 minutes and stayed there in the binary models.
The headline numbers come from the 120 to 179 minute band. Compared with zero minutes, the odds of reporting good health were 59% higher, an odds ratio of 1.59 with a 95% confidence interval of 1.31 to 1.92, and the odds of reporting high wellbeing were 23% higher, an odds ratio of 1.23 with a 95% confidence interval of 1.08 to 1.40. Read those percentages as odds, not chances. Good health was already common here, with 76.5% of the weighted sample rating their health good or very good, so a 59% rise in odds is a smaller rise in probability.
Spline models, which let duration vary smoothly instead of in 60-minute blocks, told a softer story. Gains rose up to around 120 minutes, with diminishing returns to around 200 minutes for health and 300 for wellbeing, then a flattening, with wide uncertainty beyond 400 minutes. The gains do not keep climbing with the hours.
How the 120 minutes were assembled made no difference in the study's tests. One visit of 120 minutes or more, two visits of 60 minutes or more, and three or more visits of 40 minutes or less all carried the same association, provided the weekly total crossed the line. The association held across urban and rural residents, high and low deprivation areas, men and women, people above and below 65, and people who did and did not meet physical activity guidelines. For scale, it sat in the same range as living in a low-deprivation area rather than a high-deprivation one, and meeting rather than missing activity guidelines.
The Arithmetic Nobody Did
Two hours is 120 minutes out of 10,080 minutes in a week, or 1.19% of the week, and against waking time, assuming 16 waking hours a day, it is 120 out of 6,720 minutes, or 1.79%. A factor that small leaves a lot of room for something else to be doing the work, and that smallness is part of what makes the finding surprising, and part of what makes it suspicious.
In the unweighted sample, 11,668 of 19,806 people reported zero recreational nature contact in the previous week, which is 58.9%, and nearly six in ten reported no qualifying visit at all, so the comparison driving the headline is between none and at least two hours, with the middle ground failing to separate from none in the main model.
Within the 120 to 179 minute band, 78% of unweighted observations sat at precisely 120 minutes, because people report visit lengths in round hours. Applied to that band's 1,290 people, that is about 1,006 people landing on exactly 120, and the authors flag this rounding as a reason to doubt a sharp cliff at that exact number.
The Strongest Counterargument
The strongest case against this finding starts with its design and does not need to dispute a single number, because the design cannot show that nature caused the health ratings. This is one week of self-reported visits matched to current self-rated health, in different people each wave, with no follow-up. Healthier and happier people may simply go outside more. People with flexible jobs, safe neighbourhoods, transport, company, and free time may both visit nature and rate their lives higher for reasons that have nothing to do with trees. The authors controlled for deprivation, greenspace near home, employment, and activity, and the pattern held among people with long-term illness. A control list is not a cure for reverse causality.
When health and wellbeing were modelled as ordered scales rather than binary splits, the 60 to 119 minute category did reach significance, softening the cliff into something closer to a slope. A reader who only hears "120 or nothing" has been given the binary version, not the full version, which the authors report rather than hiding.
A different dose study in the same journal points the other way on the number. Shanahan and colleagues, studying 1,538 Brisbane residents, found lower depression and high blood pressure associated with nature visits averaging at least 30 minutes, with gains to about 75 minutes for depression. The two studies measured dose differently, in different countries, with different outcomes, so they are not direct contradictions, but they show that "the dose" depends on how dose is counted.
For health, the threshold held for White British respondents but not for respondents grouped as other ethnicities. The authors suggest personal meaning, cultural practice, and a sense of place may shape what nature contact delivers. A threshold that does not travel across groups in its own sample should not be sold as universal, especially with the garden exclusion, no quality measure for the nature visited, and time in nature explaining only about 1% of variance in the unadjusted models.
What We Didn't Prove
This evidence does not prove that spending 120 minutes in nature will improve your health, because no one was assigned to a nature dose and followed over time, and it does not prove the threshold is 120 minutes rather than a softer range around two hours, that more than 300 minutes is useless, or that garden time is worthless, since garden time was excluded by definition, not tested and found wanting. It does not prove the pattern holds outside England, outside 2014 to 2016, or for groups in whom the health threshold did not hold, what kind of nature matters, or that nature itself, rather than exercise, daylight, social contact, or free time, did the work.
The Bottom Line
Two hours a week is a small, specific, and surprisingly demanding target once you count it the study's way: recreational visits to parks, woods, coasts, and countryside, not your garden and not a shopping trip, adding up across the week. In 19,806 English adults, crossing that line coincided with 59% higher odds of good health and 23% higher odds of high wellbeing, while falling short did not separate from zero in the main model, and going well beyond it added nothing in the binary analysis. Treat that as a lead worth testing in your own week, not as a prescription. The authors said it was too early for guidance, and their caveats are the reason to agree.
What You Can Do
First, count your week the study's way for one week, without changing anything yet, adding up recreational visits to parks, woodlands, beaches, and countryside, leaving out your garden and shopping trips for this count, and if your week already clears 120 minutes, you know your baseline, and if it sits at zero, like 58.9% of the unweighted sample, you know where the gap is.
Second, build the total from the splits the study actually compared: one visit of two hours, two visits of an hour, or three visits of about 40 minutes, and choose the shape that fits your week instead of forcing a shape you will abandon. A local park counts, and a short trip repeated beats a grand trip postponed, because the study's total was a weekly sum, not a wilderness badge.
Third, run a personal trial for a month and judge it by your own outcomes, not by the threshold, keeping a simple weekly note of minutes, sleep, mood, and exercise, and comparing weeks above and below 120 minutes while other habits stay roughly steady. If nothing changes, that is information too. Do not drop medical care, exercise, or sleep in favour of park time, and if mobility, safety, or access makes outdoor time hard, choose accessible, well-lit routes and company over distance.