The ten thousand step target originates from the name of a Japanese pedometer marketed in the 1960s — manpo-kei, literally "ten thousand step meter" — rather than from any study. The research that followed has consistently found that most of the mortality benefit arrives well below it, with the curve flattening somewhere between 6,000 and 8,000 steps a day for older adults.
That does not make the target harmful. It makes it a round number chosen for marketing that turned out to be in roughly the right area, which is a different claim from the one it is usually given.
What does the evidence actually show?
A dose-response curve that rises steeply and then flattens. Moving from very low activity to moderate activity produces the largest single gain; moving from moderate to high produces a smaller one.
The practical reading is that the first few thousand steps matter most, which is the opposite of how a target frames it. Someone at 2,000 steps gains far more from reaching 5,000 than someone at 8,000 gains from reaching 11,000.
Intensity appears to matter alongside volume. Steps taken at a brisk pace are associated with better outcomes than the same number taken slowly, which is why cadence has started appearing alongside step counts in the literature.
How accurate is a step count anyway?
Within about 10 per cent for walking on a wrist device, and considerably worse for anything else. Wrist-worn sensors infer steps from arm motion, so pushing a trolley, carrying a child or holding a handrail all undercount, while gesturing during a phone call overcounts.
Phone-based counting has the opposite problem: it only counts when the phone is on you. A day spent at a desk with the phone on the desk registers almost nothing regardless of how much walking happened around it.
Neither error matters for tracking a trend, which is what the number is actually good for. Both matter enormously for comparing your figure against someone else’s, or against a target set by a different device.
What does a step convert to?
Roughly 0.7 to 0.8 metres for an average adult, so 10,000 steps is about 7 to 8 kilometres. In calorie terms it is far less than people expect — the calories-burned arithmetic puts a brisk hour at a few hundred, and much of that would have been burned anyway.
| Steps | Roughly | Time at a brisk pace |
|---|---|---|
| 2,000 | 1.5 km | 15–20 min |
| 5,000 | 3.75 km | 40–50 min |
| 7,500 | 5.6 km | 60–75 min |
| 10,000 | 7.5 km | 80–100 min |
That last row is the reason the target is hard to hit incidentally. Ninety minutes of walking is not something that happens by accident in a desk-based day, which is why the number functions as a prompt to restructure the day rather than as a measurement of it.
Is a target useful at all?
Behaviourally, yes, and that is a separate question from whether it is the right number. A round, memorable, slightly ambitious figure produces more walking than an accurate personalised one that nobody remembers.
The failure mode is all-or-nothing. A target missed by a thousand steps produces the same feeling as a target missed by nine thousand, and the response to a missed target is frequently to abandon it rather than to fall back on a smaller one.
A floor works better than a ceiling for that reason: a minimum you will hit on a bad day, with the good days looking after themselves.
Substitution is where a floor earns its keep. A day that will not reach the target is still a day that can include a walk to the shop rather than a drive, and the accumulated small substitutions are what move an average over a month. Targets set at the level of the day tend to be abandoned; habits attached to existing routines are not.
What about the other round numbers?
Most of the familiar health figures have similar origins — plausible, roughly right, and far more precise than the evidence supports. Eight glasses of water a day has no clear source. Eight hours of sleep is a population average with a wide range around it. Five portions of fruit and vegetables was a communications decision rather than a threshold in the data.
None of that makes them bad advice. It makes them communication devices, and the honest framing is that they point in the right direction with a precision nobody should take literally.
Questions people ask
Is there a point where more steps stop helping? The curve flattens rather than reversing in the observational data. Very high volumes are associated with continued small benefits, and with an increased injury risk that the mortality curves do not capture.
Do stairs count for more? Yes, substantially, in energy terms — the MET arithmetic puts stair climbing several times above level walking. Most step counters do not distinguish them.
Should I aim for the target every day? Weekly totals are a more forgiving and equally valid frame. The research is generally on average daily steps rather than on consecutive days.
Does walking count as exercise? For cardiovascular and mortality outcomes, clearly. For strength and bone density it does very little, which is why guidance lists resistance work separately rather than as an alternative.
A marketing name that landed near the right answer, which is luckier than it is rigorous. The calories burned calculator puts a walk in energy terms, the TDEE calculator shows where daily movement sits in the total, and the pace calculator covers the intensity half of the question.