How long does it take to form a habit? (Not 21 days)

Where the 21-day myth comes from

In Psycho-Cybernetics (1960), Maxwell Maltz described what he saw as a plastic surgeon: patients took about three weeks to get used to a new face, amputees kept sensing a lost limb for a similar period, people needed a few weeks to feel at home in a new house. From these observations he concluded that it takes “a minimum of about 21 days” for an old self-image to dissolve and a new one to set.

Three things were lost as the idea travelled through self-help books and “21-day challenges”. The word minimum disappeared. The word about disappeared. And “getting used to a change” turned into “forming a habit”, which is a different thing: a habit is an action your context triggers without you having to decide.

None of this makes 21 days harmful as a starting point. The problem is quitting on day 22 because nothing feels automatic yet. Ben Singh and colleagues put it plainly in their 2024 review: short 21-day challenges are often not enough, especially for complex behaviours like exercise and healthy eating.

What the 66-day study actually found

The study everyone cites is Phillippa Lally’s, run at University College London and published in the European Journal of Social Psychology in 2010. It’s worth knowing the details, because the headline number hides most of the story.

• 96 volunteers each chose one eating, drinking or activity behaviour to do every day in the same context, for example “after breakfast”. Choices included drinking a bottle of water with lunch or running before dinner.
• For 12 weeks (84 days), they logged each day whether they had done it and how automatic it felt.
• 82 people provided enough data. The researchers fitted a curve to each person’s automaticity scores; it fitted for 62 of them, and fitted well for 39.
• Among those 39, the median time to reach 95% of their personal plateau was 66 days, with a range of 18 to 254 days.

Three details matter more than the median. First, the study lasted 84 days, so the longest estimates are projections of each person’s curve, not observed days. Second, the curve rises steeply at first, then flattens: early repetitions add the most automaticity. Third, missing one opportunity to perform the behaviour did not materially affect habit formation. People who performed the behaviour more consistently simply had clearer curves.

And for more than half of the 96 volunteers (57 of them), the model didn’t fit well at all. Real life is messy, and a habit doesn’t form on a schedule.

What newer research adds (2015–2025)

Keller et al., 2021 (British Journal of Health Psychology). In a randomised trial, 192 adults aged 18 to 77 each picked one nutrition behaviour and tied it either to a daily routine or to a clock time, then answered daily questionnaires for 84 days. Among those who successfully formed a habit, it took a median of 59 days to reach peak automaticity. The 2024 review that re-analysed it notes that only 23% of participants reached the study’s threshold for habit formation, with individual estimates running from 4 to 335 days. The strongest predictor was simple: repeatedly carrying out the plan after the planned cue.

Singh et al., 2024 (Healthcare). This systematic review pulled together 20 studies and 2,601 participants on physical activity, diet, water, flossing, vitamins and more. Median times ranged from 59 to 66 days and means from 106 to 154 days. Its conclusion: health habits typically take 2 to 5 months. It also found that habits practised in the morning and habits people chose themselves tended to be stronger, and that simple, repetitive behaviours with clear cues and immediate rewards (flossing, drinking water) showed larger effects than exercise or healthy eating. The authors flag the limits: few studies, and 11 of the 20 rated at high risk of bias.

Buyalskaya et al., 2023 (PNAS). Instead of asking people how automatic a behaviour felt, this team used machine learning on objective records: more than 12 million observations of gym check-ins from 30,110 members, and more than 40 million handwashing opportunities from 3,124 hospital workers. They treated a habit as formed once behaviour became highly predictable from context. Handwashing became habitual after about 9 to 10 shifts. For the gym, the original paper said 4 to 7 months, but an August 2023 correction revised it to 68 to 78 days, about 2 to 3 months. Many articles still quote “six months”; that’s the pre-correction figure.

The same data show what drives a gym habit. For 76% of members, the longer it had been since their last visit, the less likely they were to go on a given day. And 69% were more likely to go on the same days of the week they had gone before.

What “automatic” actually means

Doing something often is not the same as doing it by habit. Researchers measure habit as automaticity: the behaviour starts when the cue appears, without a decision. The standard tool is the Self-Report Behavioural Automaticity Index (SRBAI), four statements developed by Benjamin Gardner and colleagues in 2012. People rate how much they agree with each one. You can use it on yourself once a week:

A 2025 review in Trends in Cognitive Sciences (Buabang, Donegan, Rafei and Gillan) describes the brain side: a stimulus-response system that repeats well-practised actions in familiar settings, and a goal-directed system that plans. Forming a habit means shifting a behaviour from the second system to the first. That shift is gradual, which is why no single day marks the finish line.

What makes a habit form faster

The same studies point to the levers you control.

1. Make it simple. New gym members built stronger exercise habits when the routine was less complex (Kaushal & Rhodes, 2015). One push-up becomes automatic sooner than a 45-minute workout.
2. Repeat it often. A hospital worker had about 220 handwashing opportunities over 9 to 10 shifts; the average gym member in the PNAS data went every 4 to 6 days. That largely explains weeks versus months. Kaushal and Rhodes found that exercising at least four times a week for six weeks was the minimum to establish the habit.
3. Tie it to one stable cue. Same time, same place, or right after something you already do. Linking a behaviour to a routine worked as well as linking it to a clock time (Keller et al., 2021); what mattered was repeating it after the cue. Our guide to habit stacking turns this into recipes.
4. Make it pleasant. Enjoyment predicted exercise habit formation (Kaushal & Rhodes, 2015), and pleasure and intrinsic motivation strengthened the link between repetition and habit in a 16-week study of 118 people (Judah et al., 2018).
5. Choose it yourself. In one study included in Singh’s review, habits people picked themselves did better than assigned ones.
6. Consider the morning. In a small study of 48 students, a morning stretch became automatic sooner than an evening one. Our morning routine guide covers it.

What happens when you miss a day

Lally’s data are reassuring: one missed opportunity didn’t derail habit formation. The gym data add a warning: the risk grows with the gap. The longer since the last visit, the less likely the next one. So the practical rule isn’t “never miss”; it’s “don’t let one miss become three”. Plan rest days in advance so they don’t feel like failures, and do a tiny version on bad days. Our habit tracker guide covers tracking and recovering after a missed day in detail.

If you use Binome360, you can set this up in one sentence:

Frequently asked questions

Is 21 days ever enough?

Sometimes, for very simple habits with a strong cue. Lally’s fastest participant got there in 18 days, and hospital handwashing became predictable in about 9 to 10 shifts. For exercise or diet changes, plan on two months or more.

How long does it take to break a habit?

It varies just as much. In a 2025 study of 194 people trying to cut down sedentary behaviour, snacking, alcohol or smoking, the time for habit strength to stabilise ranged from 1 to 65 days (Edgren et al.). Our guide on how to break a bad habit explains the method.

Is it 66 days or 90 days?

Neither is a rule. 66 days is a median from one study; a 90-day plan is a reasonable way to cover most people, but it isn’t a research finding. Track how automatic the habit feels rather than counting down.

How do I know a habit has formed?

When you do it without deciding, and skipping it feels odd. The four SRBAI statements above are a practical weekly check.

In short

The honest answer is “about two months, often more, and it depends”: the median was 66 days in Lally’s study, and 2 to 5 months across the research. Simple, frequent, pleasant habits tied to one clear cue form fastest, and a single missed day doesn’t matter. First step: pick one small habit, attach it to something you already do every day, and rate the four automaticity statements every Sunday.

Sources

• Maxwell Maltz, Psycho-Cybernetics, Prentice-Hall, 1960.
• Phillippa Lally, Cornelia H. M. van Jaarsveld, Henry W. W. Potts and Jane Wardle, “How are habits formed: Modelling habit formation in the real world”, European Journal of Social Psychology, 40(6), 998–1009, 2010: doi.org/10.1002/ejsp.674.
• Ben Singh, Andrew Murphy, Carol Maher and Ashleigh E. Smith, “Time to Form a Habit: A Systematic Review and Meta-Analysis of Health Behaviour Habit Formation and Its Determinants”, Healthcare, 12(23), 2488, 2024: doi.org/10.3390/healthcare12232488.
• Jan Keller, Dominika Kwasnicka, Patrick Klaiber, Lena Sichert, Phillippa Lally and Lena Fleig, “Habit formation following routine-based versus time-based cue planning: A randomized controlled trial”, British Journal of Health Psychology, 26(3), 807–824, 2021: doi.org/10.1111/bjhp.12504.
• Anastasia Buyalskaya, Hung Ho, Katherine L. Milkman, Xiaomin Li, Angela L. Duckworth and Colin Camerer, “What can machine learning teach us about habit formation? Evidence from exercise and hygiene”, PNAS, 120(17), e2216115120, 2023: doi.org/10.1073/pnas.2216115120; correction in PNAS, 120(34), e2312763120, 2023: doi.org/10.1073/pnas.2312763120.
• Navin Kaushal and Ryan E. Rhodes, “Exercise habit formation in new gym members: a longitudinal study”, Journal of Behavioral Medicine, 38(4), 652–663, 2015: doi.org/10.1007/s10865-015-9640-7.
• Benjamin Gardner, Charles Abraham, Phillippa Lally and Gert-Jan de Bruijn, “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, 2012: doi.org/10.1186/1479-5868-9-102.
• Eike K. Buabang, Kelly R. Donegan, Parnian Rafei and Claire M. Gillan, “Leveraging cognitive neuroscience for making and breaking real-world habits”, Trends in Cognitive Sciences, 29(1), 41–59, 2025: doi.org/10.1016/j.tics.2024.10.006.
• Gaby Judah, Benjamin Gardner, Michael G. Kenward, Bianca DeStavola and Robert Aunger, “Exploratory study of the impact of perceived reward on habit formation”, BMC Psychology, 6, 62, 2018: doi.org/10.1186/s40359-018-0270-z.
• Marion Fournier et al., “Effects of circadian cortisol on the development of a health habit”, Health Psychology, 36(11), 1059–1064, 2017: doi.org/10.1037/hea0000510.
• Robert Edgren, Dario Baretta and Jennifer Inauen, “The temporal trajectories of habit decay in daily life: An intensive longitudinal study on four health-risk behaviors”, Applied Psychology: Health and Well-Being, 17, e12612, 2025: doi.org/10.1111/aphw.12612.

Also available in Français.