How Long a New Habit Really Takes

Why We Come Back · 12 October 2026

The 21-day rule comes from a 1960 self-help book by a plastic surgeon, not from any study of routines. Measured properly, a new habit takes anywhere from weeks to months, and the setting matters more than the calendar.

Key takeaways

The 21-day figure traces to Maxwell Maltz's 1960 book Psycho-Cybernetics, which described patients adjusting to a changed face, not people forming new routines.
In Phillippa Lally's 2010 study of 96 volunteers, the time to reach a plateau of automaticity ranged from 18 to 254 days, with 66 days usually quoted as the average.
Missing a single day did not materially affect the process in Lally's 2010 data.
In a 2011 cinema study led by David Neal, habitual popcorn eaters ate stale popcorn as readily as fresh, but only in the cinema and only with their dominant hand.
A 2023 PNAS study of 30,110 gym members and 3,124 hospital caregivers found gym routines took about four to seven months to form and handwashing weeks.

Somewhere in most self-help advice sits the same promise: do something every day for 21 days and it becomes a habit. The number is short enough to feel achievable and precise enough to sound measured. It is usually repeated without a source. It does have one, and it is not a study of behaviour at all. It comes from a book published in 1960 by a plastic surgeon, who was writing about something rather different. The research that has since tried to time the process gives a longer and much less tidy answer. It also disagrees about what to count: days on a calendar, or repetitions in a setting.

1960: A surgeon's twenty-one days

Maxwell Maltz was an American cosmetic surgeon, and Psycho-Cybernetics, published in 1960, was his book about self-image. In it he described what he had seen in his patients. People whose faces had been changed by surgery seemed to take about three weeks to get used to the new appearance. From observations like these he suggested that it takes a minimum of about 21 days for an old mental image to dissolve and a new one to settle.

Three things were lost as the idea travelled. Maltz wrote of a minimum, not an exact count. He was describing adjustment to a change, not the building of a new routine. And the figure came from clinical impressions rather than from a measured study. Writing on a university blog in 2012, the UCL researchers Benjamin Gardner and Susanne Meisel suggested the myth may come from confusing habituation, getting used to something, with forming a new behaviour.

2010: Ninety-six volunteers and a wide spread

The best-known attempt to time the process came from University College London. Phillippa Lally and colleagues asked 96 volunteers to choose an eating, drinking or exercise behaviour and to perform it once a day in the same setting, after breakfast for example, for 12 weeks. Each day they recorded whether they had done it and rated how automatic it felt on a standard questionnaire. The study appeared in the European Journal of Social Psychology in 2010.

Automaticity rose along a curve, steeply at first and then levelling off. The time people took to reach 95 per cent of their own plateau ranged from 18 days to 254, and the longest figures were projections beyond the study's 84 days. The figure usually quoted is 66 days, which the authors' later summary for family doctors gives as the average. Missing a single day did not materially set anyone back. People who performed their chosen behaviour more consistently produced curves that fitted the model better, which is consistent with steady repetition mattering more than the calendar.

The study is small, and its measure is a feeling. Of the 96 people who began, 82 provided enough data, a curve could be fitted for 62, and it fitted well for 39. So the famous number describes fewer than half of the starters, reporting how automatic a behaviour felt to them. Read generously, the study supports a measurable timescale, nearer two months than three weeks.

2011: Stale popcorn and the pull of place

The other side holds that counting days is the wrong frame. Wendy Wood and David Neal set out the case in the Journal of Consumer Psychology in 2009. On their account a consumer habit is a response cued directly by a context, such as a place, a time or a preceding action, that has repeatedly gone with it before. Such responses come to mind quickly and change slowly. Everyday pressures, including time pressure, distraction and depleted self-control, make people more likely to fall back on them. What builds the routine is repetition in a stable setting; the calendar is a by-product.

Two years later Neal, Wood and two colleagues tested this in a cinema. People who habitually ate popcorn at the cinema ate about as much stale popcorn as fresh, and were barely moved by hunger or by how much they liked it. The pattern appeared only in the cinema, and only when they ate with their dominant hand. People with weaker cinema-snacking routines ate according to how much they liked the food.

For repeat purchases and app use, this is the side that explains more. The cereal bought every week from the same shelf, or the app opened on the same morning train, is held in place by its setting more than by an opinion of it. Bas Verplanken and Wood argued in 2006 that campaigns aimed at people's beliefs rarely shift such routines, and that the best moments to intervene come when the setting changes anyway, as when people move house or change jobs. Their other suggestion works upstream, before the behaviour happens: remove the old cues and put new ones in their place.

Richard Shotton's The Choice Factory (Bias 5, pages 34 to 35) points the same way with a survey of 2,370 people: 8 per cent had switched brands without a recent life event, against 21 per cent of those who had been through one. The figures come from the book alone.

2023: Months for the gym, weeks for soap

The largest attempt to time a routine without asking anyone how it felt came in 2023. Anastasia Buyalskaya and colleagues, writing in PNAS, applied machine learning to two large records of behaviour: gym attendance for 30,110 members, with more than 12 million observations, and hospital handwashing for 3,124 caregivers, with more than 40 million. They treated a routine as formed once a person's behaviour became highly predictable from its context.

The median time to reach that point was about four to seven months for the gym, and weeks, around nine or ten shifts, for handwashing. The authors rejected the idea of a "magic number" of days. The strongest single predictor of gym attendance was the time since a person's last visit. The most predictable gymgoers also responded least to an experiment designed to get them to go more, which is what you would expect once a behaviour runs on cue.

This study measures predictability, not the felt automaticity Lally measured, and the authors note that a highly predictable gymgoer is not necessarily a habitual one. Both studies support a habit timescale of some kind, and both show it varying widely between people and between behaviours.

Both sides are measuring something real, and the second view, that timing depends on the behaviour and its setting, fits the evidence better. A timescale exists, but it belongs to a behaviour in a setting, not to people in general, and its spread is wider than any one figure can carry. Lally's own data push back: the curves did level off, and an average is a reasonable guide for planning. We would plan with it, as a range.

A timescale exists, but it belongs to a behaviour in a setting, not to people in general.

What changing a routine actually takes

Nir Eyal's Hooked, written for people who design products, sits on the same side. Its Habit Zone figure (pages 29 to 31) sets how often a behaviour occurs against how useful it seems, and the book argues that something done often enough can become automatic even when it offers little, while something done rarely may never get there. On timing, the book says that "research thus far has not found a universal timescale" (page 30), and it reports the range from Lally's study rather than the single number.

The book's frequency claim, which carries one axis of the Habit Zone, leans mainly on one exploratory study. Gaby Judah's flossing study of 50 people, published in the British Journal of Health Psychology, found that more frequent flossing predicted stronger automaticity. It also found that people told to floss after brushing their teeth, rather than before, formed stronger routines and were still flossing more often eight months later.

Put together, the research points to five practical moves for anyone changing a buying or app routine:

Plan in months. Lally's spread ran from 18 to 254 days, and gym-going took four to seven months.
Attach the new behaviour to an existing one, the way flossing after brushing outlasted flossing before it.
Keep the setting fixed: the same time, the same place, the same preceding action.
Do not restart the count after one missed day, which made no material difference in Lally's data.
To drop an old routine, change its setting: a different shop, a moved app, a new route home. Moving house or changing jobs makes this easier.

These moves come from separate studies of volunteers, gym members and hospital staff, not shoppers. They agree on the unit that matters. A new habit is built from the same action in the same place, done again until the place does the asking. Count how often you do it there, not days.

Common questions

How long does it take to form a habit?

There is no single figure. In Phillippa Lally's 2010 study the time to reach a plateau of automaticity ranged from 18 to 254 days, with 66 days usually quoted as the average. A 2023 PNAS study found gym routines took four to seven months and hospital handwashing weeks.

Where does the 21-day rule come from?

From Psycho-Cybernetics, the 1960 book by the plastic surgeon Maxwell Maltz. He observed that patients took about three weeks to adjust to a changed face, and suggested a minimum of about 21 days for a new self-image to settle. He was not describing the formation of new behaviours.

Why do people keep buying the same brand without thinking?

Wendy Wood and David Neal argued in 2009 that consumer routines are cued by context, so a familiar shelf or time of day triggers the purchase. In a 2011 cinema study by Neal, Wood and colleagues, habitual popcorn eaters ate stale popcorn as readily as fresh.

Sources

Maltz, M. (1960). Psycho-Cybernetics: A New Way to Get More Living out of Life. Prentice-Hall. Paraphrased in the article, not quoted. Full text read: not the book itself, but the 21-day passage as reported in Gardner, Lally and Wardle (2012) and Gardner and Meisel (2012), below, and as quoted in James Clear's essay (jamesclear.com/new-habit): 'a minimum of about 21 days' for an old mental image to dissolve and a new one to form.

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. Abstract read: 96 volunteers, one eating, drinking or activity behaviour daily in the same context for 12 weeks; 82 gave sufficient data; model fitted for 62, good fit for 39; time to 95 per cent of asymptote 18 to 254 days; missing one opportunity did not materially affect the process; more consistent performance associated with better fit.

Gardner, B., Lally, P., & Wardle, J. (2012). “Making health habitual: the psychology of habit-formation and general practice.” British Journal of General Practice, 62(605), 664-666. doi:10.3399/bjgp12X659466. Full text read: automaticity plateaued 'after an average of 66 days'; Maltz's patients 'typically adjusted psychologically to their new appearance within 21 days'.

Gardner, B., & Meisel, S. (2012). Blog post published 29 June 2012 on the 21-day myth, Health Behaviour Research Centre blog, University College London, blogs.ucl.ac.uk/bsh. Full text read: the myth may stem from confusing habituation with habit formation.

Wood, W., & Neal, D. T. (2009). “The habitual consumer.” Journal of Consumer Psychology, 19(4), 579-592. doi:10.1016/j.jcps.2009.08.003. Abstract read: habits are responses cued by contexts that covaried with past performance, quick to activate and slow to modify; reliance increased by time pressure, distraction and self-control depletion.

Neal, D. T., Wood, W., Wu, M., & Kurlander, D. (2011). “The pull of the past: when do habits persist despite conflict with motives?” Personality and Social Psychology Bulletin, 37(11), 1428-1437. doi:10.1177/0146167211419863. Abstract read: habitual cinema popcorn eaters ate equal amounts of stale and fresh popcorn, minimally influenced by hunger or liking, only in the cinema (Study 1) and only with the dominant hand (Study 2); weaker-habit participants ate according to liking.

Verplanken, B., & Wood, W. (2006). “Interventions to break and create consumer habits.” Journal of Public Policy and Marketing, 25(1), 90-103. doi:10.1509/jppm.25.1.90. Abstract read: belief-change interventions unlikely to change habits; downstream-plus interventions at moments such as moving household or changing jobs; upstream interventions disrupt old cues and establish new ones.

Buyalskaya, A., Ho, H., Milkman, K. L., Li, X., Duckworth, A. L., & Camerer, C. (2023). “What can machine learning teach us about habit formation? Evidence from exercise and hygiene.” Proceedings of the National Academy of Sciences, 120(17), e2216115120. doi:10.1073/pnas.2216115120. Full text read: 30,110 gym members, over 12 million observations; 3,124 caregivers, over 40 million observations; median time to habit formation 4 to 7 months for the gym and 9 to 10 shifts for handwashing; time lag since last visit the most important predictor; more predictable gymgoers responded less to the StepUp intervention; no 'magic number' of days.

Judah, G., Gardner, B., & Aunger, R. (2013). “Forming a flossing habit: an exploratory study of the psychological determinants of habit formation.” British Journal of Health Psychology, 18(2), 338-353. doi:10.1111/j.2044-8287.2012.02086.x. Abstract read: 50 participants; frequency and positive attitude predicted automaticity; flossing after brushing produced stronger habits and more frequent flossing at eight months.

Eyal, N., with Hoover, R. (2014). Hooked: How to Build Habit-Forming Products. Penguin Business, updated edition 2019. Read in full by this publication, with a reliability audit. Ch. 1, The Habit Zone, pp. 25-26 (the flossing study) and pp. 29-31 (figure 1, the frequency rule, and the line that research has not found a universal timescale).

Shotton, R. (2018). The Choice Factory: 25 Behavioural Biases That Influence What We Buy. Harriman House. Read in full by this publication, with a reliability audit. Bias 5, Habit, pp. 34-35: the author's own unpublished survey of 2,370 people on life events and brand switching.