How Long Does It Take to Form a Habit? (Not 21 Days)
Most daily habits take about two months to feel automatic — a median of 59 to 66 days across the studies that measured it — but individual people range from under three weeks to most of a year. The 21-day figure you have heard comes from a plastic surgeon's book, not from a study of habits.
That gap matters. If you expect a habit to run on its own by day 22, the long middle stretch feels like failure when it is simply the normal pace. Below is what the research found, where each number comes from, and how to plan around it. Every study is linked so you can check it.
How long does it take to form a habit?
Plan on about two months. In the best-known study, the median time for a new daily behaviour to feel automatic was 66 days, and later studies found a median of 59. The spread is wide: individual estimates run from 18 to 254 days in one study and from 4 to 335 days across a 2024 review. The kind of habit, and the person, make a large difference.
The study behind the famous 66 is Lally and colleagues, European Journal of Social Psychology, 2010. The team asked 96 volunteers to pick one eating, drinking or exercise behaviour, tie it to a daily cue such as "after breakfast", and do it every day for 84 days. Each day the volunteers rated how automatic the behaviour felt. The researchers then fitted a curve to each person's ratings and asked when it levelled off.
The answers:
- Median time to automaticity: 66 days.
- Range: 18 to 254 days. The top of that range is a projection. The study ran for 84 days, and the 254 is where one person's curve was heading, not a day anyone reached.
- About half the volunteers had not reached a habit-level score by day 84, as the British Psychological Society's summary of the study notes.
Later work lands in the same place. A randomised trial of 192 adults forming an everyday eating habit found it took a median of 59 days to reach peak automaticity among the people who formed one (Keller and colleagues, 2021). A 2024 review of 20 studies with 2,601 people found medians of 59 to 66 days, means of 106 to 154 days, and an individual range of 4 to 335 days (Singh and colleagues). The means sit well above the medians because a minority of people take a very long time.
To see those four studies drawn on one scale, with the middle marked for your kind of habit, open the habit formation calculator. It draws ranges, not a date.
Where the 21-day myth came from
The 21-day figure traces back to Psycho-Cybernetics, a 1960 book by the plastic surgeon Maxwell Maltz. He noticed that his patients took about three weeks to get used to their new faces, and wrote that it takes "a minimum of about 21 days" for an old self-image to fade. He was describing how people adjust to a change in how they look, not how they build a habit.
The quote in full: "These, and many other commonly observed phenomena, tend to show that it requires a minimum of about 21 days for an old mental image to dissolve and a new one to jell." Note the word minimum, and note what it is about.
Somewhere between 1960 and now, "at least 21 days to adjust to a new appearance" turned into "21 days to form any habit". Health researchers have tried to correct it. A 2012 paper for family doctors by Gardner, Lally and Wardle names the plastic-surgery origin, calls the 21-day idea a myth, and advises doctors to tell patients to expect a habit to take around 10 weeks.
What the habit curve looks like
A habit does not switch on at a fixed day. In the studies, automaticity rises fast at first, then each repetition adds a little less, until the curve flattens into a plateau. Most of the gain comes early; the last stretch is slow. That is why the middle weeks can feel flat even when the habit is forming exactly as it should.
Lally's team modelled each person's ratings with this kind of curve: an early climb, then a slowing, then a level line. Gardner and colleagues describe the same thing in plain terms: an "initial acceleration that slowed to a plateau". Keller's trial modelled its data with the same kind of curve.
Two practical points follow from the shape:
- The first few weeks are when repetition pays most. Each early day moves the habit further than a day in week eight.
- Slowing down is not going backwards. Around the second month, a habit often feels no easier than it did the week before. The curve predicts exactly that. It is the approach to the plateau, not a sign the habit is failing.
Older articles, including an earlier version of this one, split the process into neat phases with day counts. No study found such phases. The honest picture is one smooth curve, different for each person.
Why some habits take longer than others
More complex behaviours take longer to become automatic. In Lally's study, people who chose an exercise habit took about one and a half times as long as people who chose an eating or drinking habit. A 2024 review found morning habits and habits people chose for themselves grew stronger. The type of cue made no clear difference.
What the studies found, one factor at a time:
- Complexity. Simple actions, such as drinking a glass of water with lunch, reached their plateau sooner than routines such as 50 sit-ups before dinner (Gardner, Lally and Wardle, 2012). Exercise took about 1.5 times as long as eating or drinking habits in Lally's data, per the BPS summary.
- Time of day. In one small trial of 48 students doing a daily stretch, the group stretching after waking reached automaticity sooner than the group stretching before bed: an estimated 106 days against 154 (Fournier and colleagues, 2017). Both numbers are projections past the 90-day study, so treat it as one small finding, not a rule.
- Choice. The Singh review found that morning habits and self-chosen habits tended to be stronger. A habit you picked for yourself has that on its side.
- The cue. Keller's trial compared tying the habit to a routine ("after I brush my teeth") with tying it to a clock time ("at 8 p.m."). Both worked, and neither beat the other. What predicted the habit was doing the behaviour again and again in response to the cue.
Does missing a day reset a habit?
No. In Lally's study, missing one chance to do the behaviour did not materially affect how the habit formed, and Gardner and colleagues report that progress "soon resumed" after a single miss. The popular rule "never miss twice" is sensible advice from the writer James Clear, but no study has tested it. One missed day is a gap, not a reset.
This is one of the most useful findings in the field, because the fear of breaking a run makes people quit. The data say the curve carries on. The real risk is what a miss does to your plans: treating one missed day as proof that the whole attempt has failed, and stopping.
Lally's data were too thin to say how many misses in a row start to matter, so nobody can give you a number. The safe reading is: do it again the next day. What to do when you break a streak goes through the day after in more detail.
Is 90 days enough to form a habit?
For most daily habits, 90 days is a sensible length for a first attempt, because it covers the median of about two months with room to spare. It is not a threshold. About half of Lally's volunteers had no habit-level score at 84 days, and the average times in later studies run to 106 to 154 days. Some habits will still need work on day 91.
Ninety-day challenges are common, and the number is a reasonable planning unit. It is long enough for the curve to flatten for many people, and short enough to commit to. But you will read claims that something special happens at 90 days: that the brain "rewires", that dopamine receptors "reset", or that your identity "locks in". Those claims borrow from research on recovery from drug and alcohol addiction, which is a different process with its own timelines. No study of everyday habits has found a change at day 90.
Two ways to use a 90-day plan well:
- Count days done, not days in a row. "78 of 90 days" describes a habit that is forming. A run that resets to zero at the first gap hides that.
- Review at day 90, don't stop at day 90. Ask whether the habit now feels automatic. If it does not, that is normal; keep the cue and the habit and give it another month. The psychology of streaks explains why the count itself can start to matter more than the habit, and how to stop that.
How to plan around the real timeline
Plan for about ten weeks, not three. Tie the habit to the same daily cue, keep the first version small enough to do on a bad day, write down each day you do it, and judge your progress by how automatic it feels rather than by a date. If it still takes effort at day 60, you are in the normal range.
A short version, each point tied to its source:
- Expect about ten weeks. That is the advice Gardner, Lally and Wardle give doctors, based on the 66-day median.
- Use one fixed cue. Every study above tied the habit to a daily cue, and repeating the behaviour in response to it predicted success (Keller). Habit stacking is one way to pick the cue.
- Start small. Simpler behaviours became automatic faster (Lally; Gardner). You can build up once the small version runs on its own.
- Keep a record. A written record helps people follow through on goals; does habit tracking work? covers what 138 experiments found.
- Ask how it feels, not what day it is. The measure in these studies was a feeling: "I do it without thinking." Check that every week or two.
- Treat a miss as a gap. One missed day did not undo progress (Lally).
What the research does not say
It does not give a day on which your habit will be done. The numbers are medians and ranges from small groups of volunteers, mostly doing simple health habits, who rated their own sense of automaticity. They say nothing precise about you, about quitting a habit, or about habits far more complex than a daily walk or a glass of water.
The limits worth stating:
- Automaticity is self-reported. Every study here asked people how automatic a behaviour felt, using short questionnaires. That is a useful measure, but it is a feeling, not a scan.
- The samples are small. Lally had 96 volunteers, Keller 192, Fournier 48. The 2024 review rated most of its 20 studies at high risk of bias.
- The long tails are projections. Lally's 254 days and Fournier's 106 and 154 days extend curves past the end of the study.
- Building is not breaking. These studies measured new habits forming. Breaking an old habit, or stopping a substance, is a different question.
- No magic numbers. 21, 66 and 90 are all just numbers. 66 is a median from one study, not a finish line.
Frequently Asked Questions
How long does it take to form a habit?
About two months for a typical daily habit. Studies that tracked people forming new habits found median times of 59 to 66 days before the behaviour felt automatic. The range is wide, from under three weeks for simple actions to most of a year for harder ones, so treat two months as a planning figure, not a deadline.
Does it really take 21 days to form a habit?
No. The 21-day figure comes from a 1960 book by the plastic surgeon Maxwell Maltz, who wrote that patients needed at least 21 days to get used to a new appearance. No study of habit formation has found 21 days. The best-known study found a median of 66 days.
Is 66 days the magic number for habits?
No. 66 days is the median from one study of 96 people, and individual times ran from 18 to 254 days. A later trial found a median of 59 days. Use two months as a rough guide and judge your habit by how automatic it feels, not by the date.
How long does it take to form an exercise habit?
Longer than most simple habits. In the best-known study, people who picked an exercise habit took about one and a half times as long to reach a plateau as people who picked an eating or drinking habit. Plan on three months or more, and start with a version small enough to do on a busy day.
Is 90 days enough to form a habit?
Often, but not always. Ninety days covers the typical two months with room to spare, which makes it a good length for a first attempt. But about half the volunteers in the best-known study had not reached a habit-level score after 84 days, so some habits need longer. Nothing special happens on day 90.
Does missing a day ruin a habit?
No. In the best-known study, missing a single day did not materially affect how the habit formed, and progress resumed soon after. The popular advice not to miss two days in a row is a sensible rule of thumb, not a research finding. Pick the habit up again the next day.