How to start.
-
01
Find your real number
Track a normal week and take the average. Nearly everyone guesses high. The number you start from decides how much a change is worth.
-
02
Add one thousand
Aim for a thousand more than your average, not a round target. Each extra 1,000 a day tracked with about 15% lower mortality in the pooled data.
-
03
Use your age band
If you're 60 or over, aim at 6,000 to 8,000. Under 60, aim at 8,000 to 10,000. That's where the curve flattened.
-
04
Stop scoring the day
Missing 10,000 by 900 steps is not a failure, it's noise. Cadence barely mattered for mortality either, so don't chase speed.
Why it works.
-
Lower mortality
Across 47,471 adults with 3,013 deaths over 7.1 years median follow up, the highest step quartile (median 10,901 a day) had 53% lower all cause mortality than the lowest (median 3,553).
-
Real threshold
Benefit plateaued at roughly 6,000 to 8,000 steps a day for adults 60 and over, and 8,000 to 10,000 for adults under 60, so 10,000 is not a universal target.
-
Dose response
Across 226,889 people in 17 cohorts, each extra 1,000 steps a day tracked with 15% lower all cause mortality, and each 500 steps with 7% lower cardiovascular mortality.
Banach et al., 2023, European Journal of Preventive Cardiology
-
Low floor
The same meta-analysis found mortality benefit already detectable from about 3,867 steps a day, and cardiovascular benefit from about 2,337 steps a day.
Banach et al., 2023, European Journal of Preventive Cardiology
Who swears by it.
Steve Jobs
Held serious conversations and brainstorms on long walks, a habit his biographer Walter Isaacson documented.
Harry S. Truman
Walked a mile or two every morning at an Army cadence of 120 paces per minute.
Charles Darwin
Walked his quarter mile Sandwalk thinking path three times a day, counting laps with pebbles.
John's take.
The number is fake. Not the benefit, the number. In 1965 a Japanese company sold a pedometer called Manpo-kei, which translates to something like the 10,000 step meter, and the target has been repeated ever since by people who assume it came out of a lab. It came out of a product name. Sixty years later we’ve all been chasing a piece of Japanese ad copy on our wrists.
What the research actually says is more useful and much less flattering to anyone bragging about their step count. In the 47,471 person study, going from about 3,553 steps to 7,842 captured most of the gain, hazard ratio 0.60 down to 0.55. Pushing on to 10,901 added comparatively little. So if you’re sitting at 3,000, the next 4,000 steps are the most valuable exercise available to you and they cost you a walk. If you’re at 8,000 and grinding out laps of the kitchen at 11pm to hit a round number, you’re collecting crumbs.
Here’s my honest read. The floor matters, the ceiling doesn’t. Benefit shows up from roughly 3,900 steps. It flattens somewhere around 6,000 to 8,000 if you’re over 60 and 8,000 to 10,000 if you’re not. The two big papers don’t fully agree on whether the plateau even exists, and all of it is observational, meaning sick people walk less and that muddies everything no matter how carefully it’s adjusted for. I still walk. I just stopped treating 10,000 as pass or fail, and the day I did that was the day the habit stopped breaking.
Common questions.
How many steps a day do I actually need?
Fewer than 10,000. Mortality benefit plateaued around 6,000 to 8,000 steps a day for adults 60 and over, and 8,000 to 10,000 for adults under 60, in a study of 47,471 people. Benefit was already detectable from about 3,900 steps a day in a separate meta-analysis of 226,889 people. Pick the band for your age and stop there.
Where did the 10,000 steps figure come from?
A 1965 Japanese pedometer called Manpo-kei, literally the 10,000 step meter. It was a product name, not a research finding. Every study since has had to test the number rather than derive it, and when researchers looked, the curve flattened below 10,000 for most people.
Is 5,000 steps a day enough?
It's meaningfully better than 3,000 and a fair way short of the plateau. Mortality benefit starts around 3,867 steps a day, and each additional 1,000 tracks with about 15% lower all cause mortality. So 5,000 is real progress, not a finish line. If you're over 60, the flat part of the curve starts around 6,000.
Is walking more than 10,000 steps better?
The two big papers disagree. Paluch 2022 found benefit levelling off below 10,000, with little added between 7,842 and 10,901 steps. Banach 2023 reported no clear plateau even up to 20,000 steps, though the data at those counts are thin. Nobody has shown high counts are harmful. They just may not pay much.
Does walking faster matter?
Less than you would think. Stepping rate, meaning cadence, added little beyond total daily step count for mortality in the 47,471 person study. Total steps did the work. Walk at whatever pace you'll repeat tomorrow rather than trying to earn extra credit with speed.
Does walking actually cause longer life, or do healthy people just walk more?
This is the real weakness. All of it is observational cohort data. Reverse causation, where people who are already ill walk less, is partly adjusted for but can never be fully removed. The dose response is consistent across 226,889 people, which is suggestive. It isn't a randomized trial and shouldn't be sold as one.