Key Takeaways
  • 10,000 was never a scientific number. It most likely comes from the name of a 1965 Japanese pedometer. The research has since found the real curve
  • In 15 cohorts and 47,471 adults, mortality fell steeply and then levelled off at 6,000–8,000 steps a day for people 60 and over, and at 8,000–10,000 for younger adults
  • Most of the benefit is in the first few thousand steps above very low. A 2025 meta-analysis put 7,000 steps at 47% lower all-cause mortality than 2,000 — and 38% lower dementia and 22% fewer depressive symptoms
  • Every 1,000 extra steps was associated with about 15% lower all-cause mortality across 17 cohorts. If you're low, the next thousand is the most valuable thing on this page
  • Pace is a maybe; steps are a yes. Several large studies found brisk cadence added nothing once total steps were counted
  • A busy week isn't a lost one: hitting 8,000 on just one or two days a week captured most of the benefit of doing it daily

"Squats versus walking after meals" is the most-read article on this site, which says something about what people want from a health site: a number, and the evidence behind it. So here is the number people ask about most — how many steps a day — with the evidence behind it, including the parts that are less tidy than the step-counter on your wrist suggests.

The short answer is that it depends on your age, that it's less than you've been told, and that the most valuable steps are the first ones you add, not the last ones you chase. The longer answer is below, with every figure named to the study it came from. All of it is observational — nobody has randomised people to walk 4,000 or 9,000 steps for ten years and counted the deaths — and that caveat gets its own section, because it matters.

Where 10,000 came from

A Pedometer, Not a Study

The 10,000-step target is so familiar that it feels like it must have come from somewhere official. It didn't. Lee and colleagues, introducing their 2019 study in JAMA Internal Medicine, trace it to the trade name of a pedometer sold in Japan in 1965 by the Yamasa Clock and Instrument Company: Manpo-kei, "10,000 steps meter." The abstract of their paper puts it more bluntly — the figure has "limited scientific basis." It's a round number that happened to be printed on a gadget, and it stuck for sixty years.

That doesn't make it a bad target. For a lot of people under 60 it's roughly where the benefit stops growing. It just isn't the threshold for benefit, and treating it as one has probably discouraged more people who walk 4,000 steps a day than it has motivated.

The dose-response

The Curve: Steep, Then Flat

Evidence: Moderate — large pooled cohorts, accelerometer-measured, consistent shape

The best single dataset is Paluch and colleagues' meta-analysis in Lancet Public Health (2022), which pooled individual-level data from 15 cohorts across four continents: 47,471 adults, step counts measured by device rather than recalled, 3,013 deaths over a median of 7.1 years. They split people into quarters by daily steps and compared each with the lowest.

All-cause mortality by daily steps, Paluch 2022 Hazard ratio for death by quarter of daily steps, relative to the lowest quarter. About 3,550 steps: 1.00 (reference). About 5,800 steps: 0.60. About 7,840 steps: 0.55. About 10,900 steps: 0.47. The largest drop is between the first and second quarters. 0 0.5 1.0 Risk vs lowest group ~3,550 steps/day (lowest quarter): reference, 1.00 1.00 ~3,550 ~5,800 steps/day: hazard ratio 0.60 (95% CI 0.51–0.71) 0.60 ~5,800 ~7,840 steps/day: hazard ratio 0.55 (95% CI 0.49–0.62) 0.55 ~7,840 ~10,900 steps/day: hazard ratio 0.47 (95% CI 0.39–0.57) 0.47 ~10,900 Median daily steps in each quarter

Hazard ratio for death from any cause by quarter of daily steps, versus the lowest quarter; lines show 95% confidence intervals. Most of the drop comes from the first extra ~2,000 steps. Data: Paluch, Lancet Public Health 2022 (15 cohorts, 47,471 adults).

Median steps/dayRisk of death vs lowest95% CI
3,5531.00 (reference)—
5,8010.600.51–0.71
7,8420.550.49–0.62
10,9010.470.39–0.57

Look at where the drop is. Going from about 3,500 to about 5,800 steps a day came with 40% lower mortality. Going from 5,800 to nearly 11,000 — almost twice the extra walking — added another 13 points. The curve is steep at the bottom and flat at the top, which is the most important practical fact in this article: the person going from 3,000 to 5,000 steps is getting far more than the person going from 9,000 to 11,000.

And the flat part arrives earlier with age. In the same analysis, the benefit levelled off at 6,000 to 8,000 steps a day for adults 60 and older, and at 8,000 to 10,000 for adults under 60. Lee's study of 16,741 women with an average age of 72 (JAMA Internal Medicine, 2019) found the same shape at the older end: as few as about 4,400 steps a day was associated with 41% lower mortality than about 2,700 (hazard ratio 0.59), with the benefit levelling at around 7,500.

Under 608,000–10,000Where the mortality benefit plateaued in the pooled data (Paluch 2022).
60 and over6,000–8,000The plateau comes earlier. Above this, more steps added little.
Currently under 5,000+1,000The most valuable change on this page. Each 1,000 steps ≈ 15% lower all-cause mortality (Banach 2023).

Four more studies, same shape

The curve has been reproduced enough times, in enough populations, to be believed — even if the exact numbers differ between studies.

The 2025 update. Ding and colleagues in Lancet Public Health (2025) reviewed 57 studies from 35 cohorts and went beyond death to eight other outcomes. Against a baseline of 2,000 steps a day, 7,000 steps was associated with 47% lower all-cause mortality (hazard ratio 0.53), 25% less cardiovascular disease, 37% lower cancer mortality, 14% less type 2 diabetes, 38% less dementia, 22% fewer depressive symptoms and 28% fewer falls. The curves bent at around 5,000 to 7,000 steps. The authors graded the certainty as moderate for most outcomes — low for cardiovascular death and cancer incidence, very low for falls — and cancer incidence itself wasn't significant. Dementia rests on two studies.

How little counts. Stens and colleagues pooled 111,309 people from 12 studies in the Journal of the American College of Cardiology (2023) and asked where benefit starts. Compared with 2,000 steps a day, significant reductions appeared at about 2,500 steps for all-cause mortality and 2,700 for incident cardiovascular disease — both only just significant — and the largest reductions were at about 8,800 steps for mortality (hazard ratio 0.40) and 7,100 for cardiovascular disease (0.49).

Per thousand steps. Banach's meta-analysis of 17 cohorts and 226,889 people in the European Journal of Preventive Cardiology (2023) put it in the most usable unit: every additional 1,000 steps a day was associated with 15% lower all-cause mortality (hazard ratio 0.85), and every 500 steps with 7% lower cardiovascular mortality. (You'll see that paper quoted as saying 3,867 steps is "the minimum" — it isn't; that's just the median of its lowest group.)

Middle age, specifically. In the CARDIA cohort (Paluch, JAMA Network Open 2021), 2,110 adults with an average age of 45 wore a step counter and were followed for about 11 years. Those taking 7,000 to 9,999 steps a day had a hazard ratio of 0.28 for death compared with those under 7,000; those above 10,000, 0.45. The second number isn't better than the first — the 10,000-plus group did no better than the 7,000 group. With only 72 deaths, the confidence intervals are wide, but it's another data point for "above about 7,000, you're mostly done."

Beyond mortality

The Brain Result

The outcome people ask about second is dementia. Del Pozo Cruz and colleagues followed 78,430 UK Biobank participants aged 40 to 79 for a median of 6.9 years, with 866 dementia diagnoses (JAMA Neurology, 2022). The largest reduction in risk was at about 9,800 steps a day — a hazard ratio of 0.49, roughly half the risk. At about 3,800 steps a day the risk was already 25% lower (0.75). Those figures come from wrist-worn devices in a cohort that was 97% white, and a dementia diagnosis takes years to develop, so some people walking less were likely already in the early stages. But the direction matches the Ding meta-analysis, and it's the same curve again: a large benefit from modest walking, a larger one from more.

The fine print

Pace, Weekends, and What These Studies Can't Tell You

Evidence: Mixed for pace; limited but encouraging for weekends

Does pace matter? Less than you'd think, on current evidence. In 4,840 US adults followed for ten years (Saint-Maurice, JAMA 2020), 8,000 steps a day was associated with half the mortality of 4,000 (hazard ratio 0.49) and 12,000 with about a third (0.35) — but once total steps were accounted for, stepping faster added nothing significant (0.90). Lee's older women showed the same: intensity effects mostly vanished after adjusting for steps. Paluch's pooled data were split, with two of four pace measures still associated. Stens found a brisk cadence did help independently. The fair summary: the number of steps is the robust finding, and walking briskly is a reasonable bonus that the evidence hasn't settled.

Can you make it up on weekends? Inoue and colleagues looked at 3,101 US adults in JAMA Network Open (2023) and compared people by how many days a week they reached 8,000 steps. Those who managed it on just one or two days had a 10-year mortality risk 14.9 percentage points lower than those who never did; those who managed it on three to seven days, 16.5 points lower. The benefit plateaued at three days. That's a difference in absolute risk, not a percentage reduction, and one week of wear, but it's encouraging for anyone whose weekdays are spent at a desk.

The caveat that applies to all of it. Every study here is observational. People who walk more differ from people who walk less in ways no statistical adjustment fully captures — and, crucially, illness reduces walking. Someone with undiagnosed heart failure or early dementia walks less because they're already unwell, which makes low step counts look more dangerous than they are. Saint-Maurice's lowest group, under 4,000 steps, had a crude death rate of 76.7 per 1,000 person-years, which is partly that effect. The researchers try to handle it by excluding early deaths, but it can't be removed completely. So read the hazard ratios as the upper end of what walking does, not a guarantee. The direction, though, is consistent across every cohort, every continent and every outcome here, and it matches what the controlled exercise trials show about fitness. Walking is not a placebo.
What to do

Your Number, in Three Steps

01

Find your baseline — a week, not a day

Wear whatever you have — phone, watch, ring — for a normal week and take the average. A single day tells you nothing. If you're under 5,000, you're on the steep part of the curve and the next section is for you. If you're over 8,000 (or 6,000 and over 60), you've captured most of what steps can do, and your next gain is on the fitness and strength levels of the pyramid, not more walking.

02

Add a thousand, then another

A thousand steps is about ten minutes of walking. The easiest place to put them is right after a meal, which also blunts the blood-sugar spike — the walking-after-meals piece has the timing and the squats-versus-walking trial has the alternative if you can't get outside. Ten minutes after dinner most days moves a 4,000-step person to 5,000, which is where the curve does the most work.

03

Aim for your age's plateau, then stop counting

8,000 to 10,000 under 60, 6,000 to 8,000 over. Once you're there most days — or on three days a week, going by Inoue's data — the counter has done its job. Walk briskly when you can. And if you have the time for more exercise, the evidence says spend it on getting fitter and stronger, not on the eleventh thousand.

Common Questions
Do I really need 10,000 steps a day?
No. The figure most likely comes from the name of a 1965 Japanese pedometer, Manpo-kei — "10,000 steps meter." In the largest pooled analysis (15 cohorts, 47,471 adults) the mortality benefit levelled off at 6,000–8,000 steps for over-60s and 8,000–10,000 for younger adults. A 2025 meta-analysis found 47% lower mortality at 7,000 steps versus 2,000.
How many steps a day is good for older adults?
For people 60 and over, the pooled data plateaued at 6,000–8,000 steps a day. In 16,741 women averaging 72, about 4,400 steps was associated with 41% lower mortality than about 2,700, levelling off around 7,500.
Does walking pace matter, or just the number of steps?
Steps are the robust finding. In several large studies faster stepping added nothing once total steps were counted — in 4,840 US adults the hazard ratio for highest versus lowest cadence was 0.90, not significant — while others did find a benefit. Get the steps; walk briskly when you can.
Can I make up my steps on the weekend?
Partly. In 3,101 US adults, reaching 8,000 steps on just one or two days a week was associated with a 14.9 percentage-point lower 10-year mortality risk versus never, close to the 16.5 points for three to seven days, and the benefit plateaued at three days a week.
The Bottom Line

The steps that count most are the first ones.

Ten thousand was a pedometer's name. The research says the curve is steep and then flat: most of the benefit arrives by 6,000 to 8,000 steps if you're over 60 and 8,000 to 10,000 if you're younger, and the single most valuable change is the next thousand for someone who's walking very little. Pace is a possible bonus. Weekends count.

So measure a normal week, add ten minutes after dinner, and aim for your age's plateau — then put the rest of your effort into getting fitter and stronger, which is where the pyramid says the next years are.