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

Short answer: closer to two months than three weeks

The number 21 is catchy, but it doesn’t come from measurement, just from anecdotal observation. When psychologist Phillippa Lally’s team at University College London tracked how people pick up a new habit in everyday life, they found a median of 66 days until the automaticity of the new behavior leveled off. The median means that half of the 39 people whose results could be reliably calculated got there faster and half got there slower.

More important than the number itself is the range: from 18 to 254 days. A newer 2024 review of studies comes to a similar conclusion — expect something more like two to five months, and a lot depends on the person and on the habit.

The practical takeaway: don’t measure a habit by the calendar. Measure it by whether it already happens on its own.

Where 21 days came from

The number most likely goes back to plastic surgeon Maxwell Maltz and his 1960 book Psycho-Cybernetics. In the preface, he writes that a noticeable change in a mental image — in the book, mainly how people see themselves — usually takes at least about 21 days. As evidence, he cites everyday observations: after plastic surgery, patients take about three weeks to get used to their new face, and in his view it takes about the same time for a new house to start feeling like home.

Above all, Maltz was asking readers not to judge his exercises until three weeks had passed. It was a minimum, not a deadline, and it was about getting used to a change. He connects 21 days with habits in places like an exercise where he suggests putting on the other shoe first and tying your laces differently for 21 days — without any measurement. Writing in a journal for general practitioners, Gardner, Lally and Wardle (2012) call 21 days a myth that apparently grew out of exactly those anecdotal experiences with plastic surgery patients.

What the Lally et al. study actually found

How the study worked

Of 101 people who signed up, 96 volunteers entered the study, mostly graduate students with an average age of 27. Each chose one new behavior — usually related to eating, drinking or exercise — and was to do it every day after the same cue, such as “after breakfast.” Examples from the paper: a piece of fruit with lunch, a bottle of water with lunch, a 15-minute run before dinner.

The study ran for 12 weeks, or 84 days. Every day, participants were supposed to log on the study website whether they had done the behavior and fill in a short questionnaire on how automatically and without thinking they were doing it.

Enough data for analysis came from 82 people. A mathematical curve could be fitted for 62 of them, and it fit well for 39. Those 39 people are where the famous 66 days comes from.

What “95% of plateau” means

A habit doesn’t build evenly. At first, each repetition adds a lot; later, less and less, until the curve almost stops rising. That point is called a plateau: further repetitions don’t make the behavior any more automatic than it already is.

A curve like this only approaches its ceiling and never quite reaches it. So the authors calculated when a person reaches 95% of their maximum. On the graphs, this matched the point where the curves visibly flatten out.

Put simply: 66 days isn’t the day a habit is “formed.” It’s the time it took half of those 39 people to reach 95% of their maximum — after that, the behavior became only slightly more automatic for them.

What else the study found

• The range is huge. From 18 to 254 days. All times are calculated from the fitted curve, and since the study lasted only 84 days, anything longer than 84 days is an estimate beyond the measured data.
• Exercise had a longer median. For exercise the median was 91 days, for drinking 59 and for eating 65. The difference wasn’t statistically significant and the groups were small. The authors still take it as a hint that more complex behaviors need more time.
• Consistency matters. The curve fit better for people who did the behavior more consistently. But about half of the participants didn’t do it regularly enough for it to become a habit.
• One miss didn’t derail the process. A single missed opportunity didn’t meaningfully disrupt habit formation. What exactly happens after a slip is covered in Missed a day? Why one slip won’t break your habit.

What a newer review says

In 2024, the journal Healthcare published a systematic review with meta-analysis (Singh et al.). It covers 20 studies with 2,601 participants and habits such as physical activity, drinking water, healthier eating and flossing. But only four of them reported the time it took to form a habit — and one of those four is the Lally et al. study.

Medians in these studies ranged from 59 to 66 days, means from 106 to 154 days, and individuals from 4 to 335 days. According to the authors, health habits usually take two to five months, and short 21-day challenges probably aren’t enough. They themselves point out that there are few studies and that 11 of the 20 have a high risk of bias.

What this means for a daily walk

The study didn’t focus on walking specifically, but its design suggests a few practical points. If you’re getting started after a January resolution, see How to stick with exercise after your New Year’s resolution.

1. One fixed cue. In the study, the behavior was tied to a situation, not a clock time. The authors’ reasoning: a situation prompts the action by itself, while a time is something you have to keep an eye on. For walking, that could be “around the block after lunch” or “off the bus one stop early.”
2. The same context every time. According to the authors, habits form by repeating the same action in the same situation. So the same place and the same moment of the day are more than just a convenience.
3. A small daily bar. One that still works on days you don’t feel like it. How many steps make sense is covered in How many steps a day do you really need?
4. Expect exercise to take longer. In the study, exercise had a longer median than eating or drinking, even though the difference wasn’t statistically significant. If it doesn’t feel automatic after six weeks, that’s normal.
5. Track the feeling, not the date. Does it happen without deciding? Would it be hard not to do it? The study’s questionnaire was built on similar questions.

Why the plateau matters more than the date

A goal with a date has an end. If you set yourself “21 days” or “66 days,” you feel done after the last day — and that’s exactly when you might stop. Yet even in the study, roughly half of the well-measured people took longer than 66 days for their curve to level off.

The curve also shows that most of the change happens early. Early repetitions add the most automaticity, so the first weeks aren’t wasted, even if it doesn’t happen on its own yet. In their conclusion, the authors write that a new habit requires keeping up self-control long enough for the behavior to start running without it.

That’s why it makes sense to measure consistency rather than count down. Tada! takes a similar approach: it doesn’t count days to a finish line, it counts active days. An active day is 6,000 steps or 20 minutes of workout — whichever comes first — and your streak shows how long you’ve kept it up. It takes its data from Apple Health or Health Connect, so you can keep your watch. What exactly counts is explained in the help article What is an active day.

Habits don’t stop at 21 Tada! counts active days — 6,000 steps or 20 minutes of workout — and shows how long you’ve kept your consistency going. No sign-up, no ads. Download for iPhone

Frequently asked questions

Does 66 days apply to everyone?

No. It’s the median from 39 people, and individuals ranged from 18 to 254 days. Treat it as a rough guide, not a deadline.

Is 21 days completely useless?

It works fine as a first checkpoint. In the study, automaticity grew fastest at the start, so the first three weeks aren’t wasted. It’s just usually not the finish line.

How do I know it’s become a habit?

When you do it automatically, without thinking, and it would be hard not to do it. Participants in the study rated similar statements.

Sources

1. Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., Wardle, J. (2010, online 2009). 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
2. Maltz, M. (1960). Psycho-Cybernetics: A New Way to Get More Living Out of Life. Prentice-Hall. Open Library
3. 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
4. Singh, B., Murphy, A., Maher, C., 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. doi:10.3390/healthcare12232488
5. Tada! Help. What is an active day. tadapp.eu