10,000 steps a day started as a decades-old Japanese business slogan, not a finding. Pooled cohorts show mortality risk falling steeply up to about 6,000–8,000 steps a day for adults 60 and older, 8,000–10,000 under 60, then flattening. Steps are a floor for daily movement — not a training goal.
01Where did 10,000 steps come from?
The 10,000-step target came from a Japanese marketing campaign, not a trial — the researchers who pooled the step evidence say so on the record [2].
Not a lab. The 10,000-step goal traces to Japanese walking clubs and a business slogan that was already 30-plus years old when a 2004 review went looking for its evidence [1]. The researchers who finally pooled the step cohorts put the origin on the record: the target "is not based on evidence" — it "originates from a marketing campaign in Japan" [2]. JAMA Internal Medicine ran the drier version: the number "has limited scientific basis" [3].
It survived anyway, for the obvious reasons — it's round, it's memorable, and it shipped as the default goal in a generation of pedometers and watches. Marketing writes defaults. Evidence has to argue its way in afterward.
02How many steps a day actually matter?
Across 15 pooled cohorts, mortality risk fell steeply up to about 6,000–8,000 steps a day at 60 and older, 8,000–10,000 under 60, then flattened [2].
When the Lancet pooled 15 cohorts — 47,471 adults, median follow-up 7 years — the answer came back lower than the slogan. Cohort data, so no proof of cause; but the shape is consistent and steep exactly where the slogan says nothing [2].
Read where the gaps sit: the jump from 3,553 to 5,801 steps is worth 40 points; the jump from 7,842 to 10,901 is worth eight more [2]. In older women the same shape shows up earlier — 4,400 steps a day ran a 41% lower mortality rate than 2,700, and the curve leveled near 7,500 [3]. The expensive steps are the first ones.
By signing up you agree to receive our emails, including occasional offers, and to the privacy policy.
03Is anything special about 10,000?
No cliff at 10,000 — each extra 1,000 steps a day tracks with about 15% lower all-cause risk, with benefits detectable from 2,300–3,900 [5].
No cliff, no gate. In 78,500 UK adults wearing wrist accelerometers for a week, risk kept easing up to roughly 10,000 steps a day, and quicker half-hours added a separate sliver of benefit on top [4]. A 17-cohort meta-analysis that looked all the way out to 20,000 steps found each extra 1,000 steps a day tied to a 15% lower all-cause mortality risk, with detectable benefits starting around 2,300–3,900 [5].
Honest note: the studies disagree about the top of the curve — the Lancet pooling sees it flatten by 10,000 [2], the newer meta reads "more is better" out to 20,000 [5]. Contested ceiling, undisputed floor. Nobody's data argues about the first 8,000.
04Do people who lift need a step goal?
Steps are the floor under your training, not the training — under 5,000 a day reads sedentary on the original step index [1].
Not a goal — a floor. Training covers a few hard hours a week at most; steps measure the other 100-plus waking ones, and the cohort curves above are mostly about that background movement. The original step-index paper drew the lines that still hold up: under 5,000 a day reads sedentary, 5,000–7,499 low active, 7,500–9,999 somewhat active, 10,000 or more active [1].
So treat the count like a dashboard light, not a finish line. If the weekly average sits under 5,000, the fix is unglamorous — the school run walked instead of idled, calls taken moving, errands on foot. The twenty-minute dumbbell sessions are the training; for the rung above walking, there's zone 2, the minimum dose of easy cardio. The steps are the floor the training stands on.
Do I need 10,000 steps a day if I already lift?
Does walking speed matter, or just the count?
Are 4,000 steps a day worthless?
General-audience information — not medical advice. How we source and verify. No affiliate links in this breakdown. Last updated 2026-09-05.
Zone 2, translated: the minimum dose of easy cardio
Conversational pace, counted in minutes. The weekly target, the dose-response curve, and why your mouth beats your watch.
5 min · September 5, 2026Training20-minute dumbbell sessions that actually progress
Three moves, three days, one pair of dumbbells — and a progression rule that survives your worst week.
6 min · August 28, 2026