The science · Habits
How long does it take to form a habit? (It’s not 21 days)
What Lally and colleagues’ 2010 study of 96 people found: a median of 66 days, a range of 18 to 254, and why one missed day doesn’t matter.
The famous “21 days to form a habit” isn't a research finding — it traces back to a 1960s plastic surgeon's observation about patients adjusting to their new appearance. When researchers actually measured habit formation, the answer was slower, messier, and far more useful.
Key findings: Lally et al. (2010)
How are habits formed: Modelling habit formation in the real world. Lally, van Jaarsveld, Potts & Wardle. European Journal of Social Psychology, 40(6), 998–1009. doi:10.1002/ejsp.674
- 96 volunteers each chose an eating, drinking or activity behaviour to do daily in the same context (for example “after breakfast”) for 12 weeks; 82 gave enough data to analyse.
- Automaticity rose along an asymptotic curve. The model fitted for 62 people, 39 of them well; doing the behaviour more consistently went with a better fit.
- Time to reach 95% of each person’s automaticity plateau ranged from 18 to 254 days. The paper reports a median of 66 days.
- Missing one opportunity to perform the behaviour did not materially affect habit formation.
Want this paced for you, dated, without doing the arithmetic? Two minutes, free, no account.
What the research says
In the most-cited real-world study, Phillippa Lally and colleagues at UCL followed 96 volunteers who each chose one new daily behaviour — drinking water with lunch, a daily walk, sit-ups after coffee — and tracked how automatic it felt, day by day, for 12 weeks. Automaticity grew along a curve: fast gains early, then diminishing returns, plateauing at the point the behaviour had become habit.
- Median time to reach 95% of the automaticity plateau: 66 days — about ten weeks.
- Range: 18 to 254 days. Simple behaviours (drinking water) locked in fast; effortful ones (exercise) took months.
- Missing one opportunity did not materially affect habit formation — what mattered was returning to the routine.
The broader habit literature (notably Wendy Wood's work) points at the same engine: habits are built by repeating a behaviour in a stable context, until the context itself triggers the behaviour with little deliberation.
What this means for your plan
- Plan in ten-week horizons, not three-week ones. Anything shorter and you'll judge yourself a failure precisely when the curve says you're on track.
- Two or three new habits at a time, maximum. Each one needs daily attention until it automates; stacking six at once means none repeat reliably.
- Keep the context stable. Same trigger, same time-of-routine, same place where possible — that's the mechanism doing the work (see implementation intentions).
- Never miss twice. One miss is noise; the danger is the spiral that follows it (see why streaks break people).
How Consistency Calendar applies this
Sprints run for weeks to months by design — long enough for the automaticity curve to do its work. The plan validator caps habits and warns when more than three start on day one, nudging extra habits to phase boundaries instead. And because Lally found single misses harmless, your headline score is a rolling 28-day execution percentage, not an unbroken chain.
Build a habit that lasts ten weeks →or connect it to Claude or ChatGPT
Quick answers
How long does it take to form a habit? In Lally et al. (2010), the time to reach 95% of each person’s automaticity plateau ranged from 18 to 254 days, with a median of 66 days — about ten weeks. It depends heavily on the person and the behaviour.
Is it really 21 days to form a habit? No. The 21-day figure is not a research finding; when habit formation was actually measured, it took a median of 66 days.
Does missing a day break a habit? No. In Lally et al. (2010), missing one opportunity to perform the behaviour did not materially affect the habit formation process. What matters is returning to the routine.
How many people were in the Lally habit study? 96 volunteers took part for 12 weeks; 82 provided enough data for analysis, and the habit curve was modelled for 62 of them.
What is the full citation and DOI? Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009. doi:10.1002/ejsp.674
Key references
- Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6).
- Wood, W., & Neal, D. T. (2007). A new look at habits and the habit–goal interface. Psychological Review, 114(4).
- Maltz, M. (1960). Psycho-Cybernetics — the anecdotal origin of the “21 days” myth.