The blog

How long does it take to form a habit? What the studies actually measured

A person writing notes beside a laptop at a home desk

It is day 22. Three weeks ago you decided to open each workday by writing a three-line plan before touching your inbox. You have done it most mornings. It still takes a small push every time, and on Tuesday you forgot until lunch. Somewhere you read that a habit takes 21 days, so the obvious conclusion is that it has not worked.

That conclusion is wrong, and the number behind it was never a research finding. When researchers have followed people building a new daily behaviour, the typical time for it to start feeling automatic has been closer to two or three months, with some people getting there in a few weeks and others still climbing after the study ended. Just as important, those numbers describe something more specific than most summaries admit: the point where a self-rated feeling of "I do this without thinking" stops rising.

Short answer: there is no fixed number of days. In the study most often quoted, the median was 66 days and individual estimates ran from 18 to 254 days. A 2024 systematic review of 20 studies found medians of 59 to 66 days and means of 106 to 154 days, with individuals ranging from 4 to 335 days, and concluded that most health habits need about two to five months. Simple actions done every day in the same setting settle fastest. Missing a single day did not derail the process in the original study. All of these figures come from health behaviours such as drinking water, eating fruit, stretching and exercise. We found no study that timed how long a work or study habit takes to plateau.

Where the 21-day figure came from

The 21-day claim is usually traced to Psycho-Cybernetics, a 1960 book by the plastic surgeon Maxwell Maltz. A commentary written for family doctors by Gardner, Lally and Wardle (2012) describes the origin plainly: the figure appears to come from anecdotal observations that patients adjusted psychologically to a changed appearance within about 21 days. That is an observation about getting used to something. It says nothing about repeating a behaviour until it runs on its own, and no experiment stood behind it.

The same authors warn about why the myth matters. Someone who expects a habit in three weeks is likely to quit during week four, at exactly the stage where the behaviour is still becoming easier.

What "formed" means in the research

Before any number makes sense, you need to know what was being counted. In psychology, a habit is not simply something you do often. It is an action that a familiar setting triggers with little deliberate thought: you sit down in the car and your hand reaches for the seatbelt.

Researchers call that quality automaticity, and in most habit studies it is measured by asking people. The Self-Report Habit Index and its short form, the Self-Report Behavioural Automaticity Index, ask how much you agree with statements about a behaviour. The four items of the short form, validated by Gardner and colleagues (2012), describe a behaviour as something "I do automatically", "I do without having to consciously remember", "I do without thinking" and "I start doing before I realize I'm doing it".

When participants answer those questions day after day, their scores tend to rise quickly at first and then level off. Researchers fit a curve to each person's scores and define the habit as "formed" at the day the curve reaches 95 percent of its own ceiling. Three things follow from that definition:

  • It is a plateau, not a finish line. The number marks when the feeling of automaticity stops growing for that person, not when the behaviour becomes certain.
  • The ceiling differs between people. One person may level off at "almost fully automatic" and another at "somewhat automatic". Both count as reaching their plateau.
  • It rests on self-report. People are rating their own inner experience, which is useful but imperfect.

The 66-day study, read closely

The figure that replaced 21 days in most articles comes from Lally, van Jaarsveld, Potts and Wardle, published in the European Journal of Social Psychology (online in 2009, in the 2010 volume). The design was simple. Ninety-six volunteers each chose one eating, drinking or activity behaviour to carry out once a day in the same context, for example after breakfast, for 12 weeks. Every day they recorded whether they had done it and completed the habit index.

The details are worth knowing, because they are rarely repeated:

  • Of the 96 volunteers, 82 provided enough data to analyse.
  • The expected rise-then-plateau curve could be fitted for 62 of them, and it was a good fit for only 39.
  • For those people, the estimated time to reach 95 percent of their plateau ranged from 18 to 254 days. The 2024 review cited below reports the median as 66 days.
  • The study itself ran for 84 days. Any estimate longer than that is a projection from the fitted curve, not something the researchers watched happen.
  • Performing the behaviour more consistently was associated with a better fit to the curve.

So "66 days" is the middle of a wide spread, drawn from the participants whose data followed the expected pattern, in one study of simple health behaviours. The authors' own summary stressed the variation and that habit formation "can take a very long time". The Gardner commentary adds that automaticity peaked more quickly for simple actions such as drinking water than for more elaborate routines such as doing 50 sit-ups, and suggests telling people to expect about 10 weeks of daily repetition.

What later studies found

Several research groups have since timed habit formation in other ways. The table lists the studies that reported a duration.

StudyWho and whatHow habit was measuredTime reported
Lally and colleagues (2010)96 volunteers, a chosen daily eating, drinking or activity behaviour, 12 weeksDaily self-reportMedian 66 days, range 18 to 254
Kaushal and Rhodes (2015)111 new gym members, 12 weeksSurveysAt least four sessions a week for about 6 weeks as the minimum for an exercise habit
Fournier and colleagues (2017)48 students, a daily stretch, 90 daysDaily self-report in an appAbout 106 days when done on waking, about 154 days when done before bed (extrapolated)
Keller and colleagues (2021)192 adults, a chosen everyday nutrition behaviour, 84 daysDaily self-reportMedian 59 days among those who successfully formed the habit
Buyalskaya and colleagues (2023, corrected)30,110 gym members and 3,124 hospital workersRecorded behaviour, scored for predictabilityAbout 2 to 3 months for gym visits, a couple of weeks for handwashing at work

The randomised trial by Keller and colleagues (2021) is the closest thing to a repeat of the original design. Its median of 59 days sits close to 66. It also carries a sobering detail that the 2024 review points out: only 23 percent of participants reached the study's threshold for a formed habit, and individual times ranged from 4 to 335 days. The median describes the people for whom it worked.

Fournier and colleagues (2017) assigned students to do a hip stretch either on waking or before bed. The morning group was estimated to reach automaticity about seven weeks sooner. Both figures are extrapolations from learning curves, and the evening figure lies beyond the 90 days the students were actually followed.

Kaushal and Rhodes (2015) followed new gym members and estimated that exercising at least four times a week for six weeks was the minimum needed to establish an exercise habit. Consistency, low complexity, the exercise environment and positive feelings about the sessions each predicted stronger habits.

The systematic review and meta-analysis by Singh and colleagues (2024) pulled this literature together: 20 studies and 2,601 participants. Only four of the 20 reported a time to habit formation. Those gave medians of 59 to 66 days and means of 106 to 154 days, and the authors concluded that habits can start forming within about two months and that most health behaviours need roughly two to five months to become automatic. They also rated 11 of the 20 studies as having a high risk of bias, and the pooled results varied a great deal from study to study.

A 2026 study adds a caution. McAlpine and colleagues tracked a physical activity habit in 41 adults every two days for 12 weeks during the 2020 lockdowns. Habit strength rose overall, but the growth varied so much between people that neither a straight line nor the expected flattening curve fitted the group well. The authors note that the original curve model has never been directly replicated. In their data, people with higher self-control built habit strength faster.

The gym figure that was corrected

One study avoided self-report altogether. Buyalskaya and colleagues (2023) used machine learning on two very large sets of recorded behaviour: gym check-ins from 30,110 members, with a median of more than four years of data each, and electronically monitored handwashing by 3,124 hospital workers. Instead of asking people how automatic a behaviour felt, they measured how predictable each person's behaviour became from context, such as the day of the week and the time since the last visit, and looked for the point where that predictability levelled off.

Their headline finding stands: there is no "magic number". Handwashing at work became predictable within a couple of weeks, after about 9 to 10 shifts. Going to the gym took months.

How many months is where care is needed. The article as first published in April 2023 gave an estimate of 122 to 226 days, about four to seven months, for gym habits. In August 2023 the journal published a correction: the estimate should read 68 to 78 days, about two to three months, and a figure and a supplementary table were replaced. Two widely read explainers we inspected on 13 November 2026 still describe the gym result as about six months. The corrected figure brings this objective measure much closer to the self-report studies above.

The gap between handwashing and the gym still makes sense. Hospital workers meet dozens of handwashing opportunities per shift in the same setting. A gym visit happens far less often, takes planning, and competes with everything else in the day. The authors make the same point: gym attendance is a less frequent and more complex behaviour.

Why the range is so wide

Across these studies, a handful of factors keep appearing.

  • How simple the action is. A glass of water levels off sooner than a workout. Low complexity predicted stronger habits among the new gym members.
  • How often and how consistently you repeat it. Repetition was the key predictor of automaticity in the Keller trial, and consistency mattered in the original study.
  • How stable the setting is. Habits are tied to cues. In a study of transferring university students, Wood, Tam and Witt (2005) found that exercise, reading and television habits survived the move only when parts of the usual context stayed the same.
  • Time of day. One small study favoured mornings. The 2024 review lists timing among the influences, on limited evidence.
  • Whether you like doing it. In a study of flossing and a daily vitamin tablet, Judah and colleagues (2018) found that pleasure and intrinsic motivation strengthened the link between repetition and habit.
  • The person. Even with the same behaviour and the same instructions, individuals differ several times over, and the reasons are not well understood. Trait self-control is one candidate, but the evidence conflicts: it went with faster habit growth in the 2026 study above and made no detectable difference in a 90-day study from 2020.

What about work and study habits?

This is the question most readers of a focus blog actually have, and the direct evidence is thin. Every duration in the table above comes from a health behaviour. In our PubMed searches on 13 November 2026 we did not find a study that estimated how many days a work habit, such as a daily planning ritual or a fixed deep work block, takes to reach its plateau.

There are some useful neighbours:

  • Study habits respond to a stable context. Stojanovic, Grund and Fries (2022) had 95 university students build a new study habit over six weeks. Half were told to keep the place and time stable and half to keep changing rooms and times. A stable context predicted more automaticity and better completion of each session's goal. The same pattern appeared in data from 218 users of a habit app. The study did not estimate a time to plateau, and its participants were mostly young, with most of the students rewarded with course credit.
  • Workdays are natural habit anchors. In interviews with ten people in a habit-based programme, Lally, Wardle and Gardner (2011) found that habits were typically formed in work-based contexts. Weekends and holidays temporarily disrupted them because the cues were missing, and the habits came back on returning to work. New behaviours felt effortful at first and easier as automaticity grew.
  • Non-health habits follow a similar early pattern. A 90-day field study in which adults chose behaviours ranging from eating fruit to saving money or being more patient found that habit strength rose substantially over three months, especially for people who performed the behaviour consistently (van der Weiden and colleagues, 2020).

A reasonable reading is that a small, daily, well-anchored work habit probably behaves like the simpler health habits, and a long or demanding one like the gym. That is an inference from neighbouring research, not a measured result.

Does missing a day reset the clock?

No, on the evidence available. In the original study, missing one opportunity to perform the behaviour "did not materially affect the habit formation process". The Gardner commentary describes automaticity gains as resuming soon after one missed performance.

That finding has limits. It concerns a single missed opportunity, not a missed week, and the same study linked more consistent performance to a cleaner habit curve. The practical message is to count a missed day as noise and get back to the cue the next time it comes round. Streak counters that return to zero after one lapse are stricter than the research supports.

A practical way to use this

  1. Budget about ten weeks, then review. Plan for two to three months of daily repetition for a simple habit, longer for a demanding one. Do not judge it at day 21.
  2. Make the action small and specific. "Write three lines of plan" will settle faster than "plan my day properly".
  3. Attach it to a cue you meet every day. Linking the action to an existing routine or to a set time worked equally well in the Keller trial. An if-then plan is a tidy way to write the link down, and our guide to implementation intentions shows how. If you plan your day in blocks, give the habit a fixed slot, as described in our article on time blocking.
  4. Keep the setting the same. Same desk, same point in the morning, same first step. A constantly changing context went with weaker study habits in the experiment above.
  5. Consider the morning, if it suits you. The evidence is one small study, so weigh it against your own rhythm. Our article on chronotype covers how time-of-day preference varies between people.
  6. Choose something you do not dread. Enjoyment helped repetition turn into habit.
  7. Expect weekends and travel to interrupt. Restart at the next normal cue.
  8. Judge by effort, not by the calendar. The useful sign is that starting takes less deliberate thought than it did a month ago.

Habits also earn their keep on tired days. Across five studies, Neal, Wood and Drolet (2013) found that when people's self-control was low they fell back on their existing habits, helpful and unhelpful alike. A work routine that has become automatic is one you are more likely to keep on the days when willpower is short. Our article on decision fatigue looks at what that research does and does not show.

What the research cannot tell you

  • The samples are small and the behaviours narrow. Four studies in the 2024 review reported a time to habit formation, all for health behaviours.
  • Most figures rest on self-report and on curves that do not fit everyone. In the original study, fewer than half of the analysed participants showed a good fit.
  • Long estimates are projections. Figures such as 154 or 254 days extend past the end of the studies that produced them.
  • Not everyone gets there. In one trial, about a quarter of participants reached the threshold within 84 days.
  • The core model has not been directly replicated, and a 2026 attempt to fit it to group data did not succeed.
  • Breaking a habit is a different question. None of these studies timed how long an unwanted habit takes to fade.

Where Zone fits

Zone is a pair of earbuds with in-ear EEG and an app. None of the studies in this article used Zone, and Zone does not build habits for you or shorten the time they take.

Its possible role is narrower. A habit attaches to a point in your day, and choosing that point is easier when you know how your attention usually behaves. Our guide to focus tracking describes the daily patterns Zone is built to show, such as when in the day your focus tends to be strongest and how long you usually last before attention drops. The app presents these as five focus states, which are app brain states: estimates defined by the app, not laboratory measurements. Use them as one input, next to your own experience, when you decide where a new work habit should sit. Zone is a wellness and performance tool and is not medical care.

Common questions

How long does it take to form a habit?

There is no single number. Studies of daily health behaviours report medians of about 59 to 66 days, with individuals ranging from a few days to most of a year. A 2024 review concluded that two to five months is typical.

Is the 21-day rule true?

No. It traces back to a 1960 book describing how long patients took to get used to a changed appearance, not to research on repeated behaviour. Studies that tracked habits found longer and far more variable times.

Where does 66 days come from?

From a study published in the European Journal of Social Psychology in which volunteers repeated a chosen daily behaviour for 12 weeks. It is the median estimate, within a range of 18 to 254 days.

How many days does it take to build a simple habit?

Simple actions done daily in a fixed setting, such as drinking a glass of water after breakfast, reached their plateau sooner than effortful routines. The shortest estimate in the original study was 18 days.

Does missing one day ruin a habit?

The original study found that missing one opportunity did not materially affect habit formation. Repeated gaps are a different matter, because consistency predicts stronger habits.

How long does it take for exercise to become a habit?

One study of new gym members suggested at least four sessions a week for about six weeks as a minimum. A large analysis of gym check-ins, after correction, estimated about two to three months.

Do habits form faster in the morning?

One study of 48 students found that a stretch done on waking was estimated to become automatic sooner than the same stretch done before bed. It is a single small study, so take it as a hint.

How long does it take to break a habit?

The studies here timed forming habits, not breaking them. Other research shows that habits weaken when the context that cues them changes, but it does not give a number of days.

How long does it take to form a work or study habit?

We found no study that measured this directly. A six-week experiment showed that study habits became more automatic when the place and time were kept stable, but it did not estimate a time to plateau.

Can Zone tell me when a habit has formed?

No. Zone shows app-defined estimates of focus over time. It does not measure habit strength.

Sources

  1. Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., and Wardle, J. (2010). How are habits formed: modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998 to 1009. https://doi.org/10.1002/ejsp.674
  2. Singh, B., Murphy, A., Maher, C., and Smith, A. E. (2024). Time to form a habit: a systematic review and meta-analysis of health behaviour habit formation and its determinants. Healthcare, 12(23), 2488. PubMed record
  3. Gardner, B., Lally, P., and Wardle, J. (2012). Making health habitual: the psychology of habit-formation and general practice. British Journal of General Practice, 62(605), 664 to 666. PubMed record
  4. Gardner, B., Abraham, C., Lally, P., and de Bruijn, G. J. (2012). Towards parsimony in habit measurement: testing the convergent and predictive validity of an automaticity subscale of the Self-Report Habit Index. International Journal of Behavioral Nutrition and Physical Activity, 9, 102. PubMed record
  5. Keller, J., and colleagues (2021). Habit formation following routine-based versus time-based cue planning: a randomized controlled trial. British Journal of Health Psychology, 26(3), 807 to 824. PubMed record
  6. Fournier, M., and colleagues (2017). Effects of circadian cortisol on the development of a health habit. Health Psychology, 36(11), 1059 to 1064. PubMed record
  7. Kaushal, N., and Rhodes, R. E. (2015). Exercise habit formation in new gym members: a longitudinal study. Journal of Behavioral Medicine, 38(4), 652 to 663. PubMed record
  8. Buyalskaya, A., and colleagues (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. PubMed record
  9. Correction for Buyalskaya and colleagues (2023). Proceedings of the National Academy of Sciences, 120(34), e2312763120. PubMed record
  10. McAlpine, T., and colleagues (2026). A longitudinal study examining physical activity habit formation. Behavioral Sciences, 16(4), 535. PubMed record
  11. Stojanovic, M., Grund, A., and Fries, S. (2022). Context stability in habit building increases automaticity and goal attainment. Frontiers in Psychology, 13, 883795. PubMed record
  12. Lally, P., Wardle, J., and Gardner, B. (2011). Experiences of habit formation: a qualitative study. Psychology, Health and Medicine, 16(4), 484 to 489. PubMed record
  13. van der Weiden, A., and colleagues (2020). How to form good habits? A longitudinal field study on the role of self-control in habit formation. Frontiers in Psychology, 11, 560. PubMed record
  14. Judah, G., and colleagues (2018). Exploratory study of the impact of perceived reward on habit formation. BMC Psychology, 6(1), 62. PubMed record
  15. Wood, W., Tam, L., and Witt, M. G. (2005). Changing circumstances, disrupting habits. Journal of Personality and Social Psychology, 88(6), 918 to 933. PubMed record
  16. Neal, D. T., Wood, W., and Drolet, A. (2013). How do people adhere to goals when willpower is low? The profits (and pitfalls) of strong habits. Journal of Personality and Social Psychology, 104(6), 959 to 975. PubMed record

Related reading

Read your own focus.

Zone turns focus into a live reading, from inside your ears.

Reserve early access

← All articles