How Long Does It Take to Form a Habit? What Studies Found

In the studies that timed it, a new daily habit took a median of about two months to become automatic: 59 days in one study and 66 days in another. Where a study reported a mean, it was longer, from 91 to 154 days. Individual people ranged from 4 days to 335 days. The 2024 review that collected these studies advises expecting two to five months, and says 21 days is not enough.

This page goes through what each study measured, who took part, and what the numbers can and can't tell you. The figures come from the 2010 paper behind the 66-day number and from a 2024 systematic review, both read in full on 8 October 2026.

The four studies that timed a habit

Study and habit	Result
Lally and others, 2010: one daily eating, drinking or exercise behaviour, chosen by each person	Median 66 days to reach 95% of the person's own plateau; range 18 to 254
Keller and others: healthy eating	Median 59 days to reach a set habit threshold; range 4 to 335
Lally and others, an earlier study: healthy eating and self-weighing	Mean 91 days, as people reported it themselves
Fournier and others: daily stretching	Mean 106 days in the morning group, 154 days in the evening group

The last three rows are given as the 2024 review reports them (Singh and others). We read the review in full, but we did not open those three papers. The review covered 20 studies with 2,601 participants, and only these four reported a time.

In the healthy-eating study with the 59-day median, only 23% of the participants reached the threshold at all. So the median describes the people who got there, and most did not within the study.

Inside the 66-day study

66 days is the number most people quote, so it helps to know where it comes from.

• Who took part. 96 volunteers, recruited through a university email: 30 men and 66 women, mostly postgraduate students, with a mean age of 27.
• What they did. Each person chose one behaviour they wanted to turn into a habit: 27 chose something to eat, 31 something to drink (most often water), 34 an exercise and 4 something else. They tried to do it every day for 84 days, in the same situation each day, such as after breakfast.
• How it was measured. Each day they rated how automatic the behaviour felt, on a questionnaire scored from 0 to 42. The researchers fitted a curve to each person's scores and counted the days until the curve reached 95% of that person's own plateau.
• How many people produced a number. 14 dropped out. The other 82 logged on for a median of 47 of the 84 days. The curve could be fitted for 62 of them, and it was a good fit for 39, which is 48% of the 82.

The 66 days is the median of those 39 people. Half of them fell between 39 and 102 days. The shortest time was 18 days and the longest 254. The study ran for 84 days, so the longer times are read off each person's fitted curve and were not observed.

The 39 were the consistent ones: they did their behaviour on a median of 88% of the days they reported. The authors write that even among these motivated volunteers, "approximately half did not perform the behaviour consistently enough to achieve habit status".

So the careful version of the claim is this: in one study, 39 volunteers who kept at a simple daily behaviour took a median of 66 days for it to feel about as automatic as it was going to get, and some needed more than three times as long.

Does the kind of habit matter?

Probably, though the evidence is thin.

In the 2010 study the median was 65 days for eating behaviours (10 people), 59 days for drinking (15 people) and 91 days for exercise (13 people). That difference was not statistically significant, and the authors say the study was too small to settle it. They also note that the exercise group took about one and a half times as long as the other two.

For exercise there is one more number. In a study of 111 new gym members, exercising at least four times a week for six weeks was the minimum needed to form an exercise habit (Kaushal and Rhodes, 2015).

The 2024 review also pooled 12 studies that measured habit strength before and after a programme. Scores rose by a standardised mean difference of 0.69 overall: 1.11 for flossing, 0.69 for physical activity and 0.57 for diet. By the usual convention 0.2 is a small difference, 0.5 a medium one and 0.8 a large one. These compare scores before and after a programme, so they show that habit scores went up and can't show what caused it.

The review picked out a few things that went with stronger habits:

• Time of day. In the stretching study, morning stretchers took a mean of 106 days and evening stretchers 154.
• Choosing it yourself. Habits people chose for themselves grew stronger than habits they were assigned.
• Simple behaviours. Drinking water became a stronger habit than exercise or eating fruit and vegetables.
• The type of cue made no clear difference. One study compared a time cue (a glass of water at a set time) with a routine cue (a glass of water after breakfast) and found no significant difference.

Does missing a day reset a habit?

Not in the 2010 study. The researchers found 140 single missed days across 55 people. On the day after a miss, the automaticity score was lower by an average of 0.29 points on the 0 to 42 scale, which they call "a very small decrease".

They also compared two-day stretches. When the behaviour was done on both days, the score rose by a mean of 0.79 points. When there was a missed day in between, it rose by 0.55. Their conclusion is that a missed opportunity "did not materially affect the habit formation process".

A missed week is a different matter. The authors point to an earlier study of exercise, which we did not open, and sum up the two together: "missing one opportunity does not preclude habit formation, but missing a week's worth of opportunities reduces the likelihood of future performance and hinders habit acquisition".

If you track a habit as a streak, this is the case for a rule that gets you back the next day. How to build a daily streak has one.

Where does 21 days come from?

The 2024 review traces the idea to Maxwell Maltz's book Psycho-Cybernetics. We have not opened the book, so we can't say what it claims in its own words. We found no study that measured 21 days.

The review's own advice is that people "should anticipate a timeframe of at least two to five months". It adds: "Brief 21-day challenges or kickstarts are unlikely to be sufficient to firmly ingrain new habits."

What these numbers can't tell you

• The samples are small and particular. The 66 days rests on 39 people, mostly postgraduate students who volunteered.
• "Automatic" is self-rated. The 2010 study used a daily questionnaire, and the authors note it had never been used day after day before.
• The yardsticks differ. One study counted days to 95% of a person's own plateau, another counted days to a fixed threshold, and a third asked people how long it took. The results sit side by side in the table above, but they don't measure the same thing.
• The wider evidence is weak. Of the 20 studies in the review, 11 were rated at high risk of bias and 5 at low risk. A third had no control group. The studies had between 20 and 537 participants, with a median of 103.
• It covers health habits only. The studies are about eating, drinking water, exercise, flossing and similar behaviours. The review searched for studies up to 18 October 2023.
• Two values for one analysis. The review's results section gives the pooled difference as 0.69 and its discussion gives 0.67. We use the results section.
• Two to five months is advice. It is the review authors' reading of four studies, and no study measured it as such.

What to do with this

1. Plan for months. Two to five months is a fairer expectation than three weeks. A 30-day sheet is the first of several, so print the 30-day habit tracker again when the first one is full.
2. Tie the habit to something you already do every day. The volunteers in the 2010 study did their behaviour in the same situation each day, for example after breakfast.
3. Pick it yourself and keep it simple. In these studies drinking water became a stronger habit than exercise did. These habit ideas with a daily minimum are sized that way.
4. Do it in the morning if you have the choice. This rests on one study of stretching, so treat it as a tiebreaker.
5. After a missed day, carry on the next day. One miss cost almost nothing in the study. A missed week is what to avoid.

For more figures on habits, goals and New Year's resolutions, each with its source, see the resolution and habit statistics. To keep a habit next to the goal it serves, GoalTracker holds your goals and their steps in your browser, free and with no account.

Sources

1. Lally, van Jaarsveld, Potts and Wardle (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. Read in full.
2. Singh, Murphy, Maher and Smith (2024). Time to Form a Habit: A Systematic Review and Meta-Analysis of Health Behaviour Habit Formation and Its Determinants. Healthcare 12(23), 2488. doi:10.3390/healthcare12232488. Open access, read in full.
3. Kaushal and Rhodes (2015). Exercise habit formation in new gym members: a longitudinal study. Journal of Behavioral Medicine 38(4), 652-663. doi:10.1007/s10865-015-9640-7. Abstract only.

FAQ

How long does it take to form a habit?

About two months at the median in the studies that timed it: 59 days in one and 66 days in another, with individual people ranging from 4 to 335 days. A 2024 review of 20 studies advises expecting two to five months.

Is it true that it takes 21 days to form a habit?

No study we found measured 21 days. The 2024 review traces the idea to the book Psycho-Cybernetics and says 21-day challenges are unlikely to be enough.

Does it take 66 days to form a habit?

66 days is a median from one study. It describes 39 volunteers whose data fitted the researchers' model, and their times ranged from 18 to 254 days. It is a midpoint for that group and was never meant as a rule.

Does missing one day ruin a habit?

Not in the study that tested it. After a single missed day the automaticity score was 0.29 points lower on a 0 to 42 scale, and the authors concluded that a miss did not materially affect habit formation. They treat a missed week as a different matter.

How long does it take to form an exercise habit?

In one study of 111 new gym members, the minimum was at least four sessions a week for six weeks. In the 2010 study the median for exercise behaviours was 91 days, against 59 for drinking and 65 for eating, though that difference was not statistically significant.

What makes a habit form faster?

The 2024 review found stronger habits for behaviours people chose themselves, for simple behaviours such as drinking water, and for a stretching routine done in the morning. Each of these rests on one or two small studies.