You have probably read it somewhere: do something for 21 days and it becomes a habit. So you try: a glass of water after breakfast, ten pages before bed, a short walk. Around day 12 it feels like a chore, and on day 21 you are still talking yourself into it.
Then comes the quiet conclusion: I'm just bad at habits. It's the wrong conclusion. The number was off, not you. Here is where "21 days" came from, what researchers have measured instead, and how to build a habit when you know it will take longer.
Where "21 days" came from
In 2012, Benjamin Gardner, Phillippa Lally and Jane Wardle wrote about habit formation in the British Journal of General Practice. They call the 21-day figure a myth. It traces back to an anecdote from a plastic surgeon, Maxwell Maltz.
A good story is not a measurement. Researchers have measured how long habits take, and the answer is longer and messier than 21 days.
What the research found
Phillippa Lally and her colleagues (2010) followed people building a new daily habit. The habit grew more automatic, meaning it took less effort and less arguing with yourself, until it leveled off after about 66 days on average. The time differed widely from person to person. And missing a single day did not seriously harm the process.
A 2024 systematic review by Singh and colleagues pulled together 20 studies with 2,601 people in total. The median time to form a habit came out at about 59 to 66 days. Individual times went up to 335 days, which is close to a year.
So roughly two months is a typical figure, not a deadline, and for some people it is much longer. Neither study can tell you your own number.
Why the number varies so much
In a diary study, Wendy Wood, Jeffrey Quinn and Deborah Kashy (2002) found that roughly 35 to 43% of everyday actions were habits, done in the same place almost every day. That is a big share of an ordinary day already running on routine, and a new habit has to find its place in it.
The studies above show that the time varies. They don't tell you why for any one habit. What follows is our reasoning, not a finding:
- A bigger action asks more of you each day, so there are more ways for a day to go wrong. A glass of water has almost none. A 30-minute workout has plenty: time, energy, weather.
- A cue that comes at the same time and in the same place gives the habit somewhere to live. That matches what the diary study counted as habits.
- A cue that moves around, like "sometime after work", gives it nowhere to sit.
Step 1. Choose one habit and write it as one sentence
Start with one. Write what you will do and when, in a single sentence: "After breakfast, I drink a glass of water." Skip "be more active" or "read more". Nothing in them tells you what to do at 8:15 on a Tuesday.
Phrase it as something to start, not something to stop. Why that matters is in our guide on how to keep your New Year's resolutions past January.
Step 2. Pick a cue that already happens every day
A cue is the moment that tells you: now. The easiest to follow is one that already happens at the same time and in the same place, like the kettle clicking off or parking the car. It works for any area of life:
- Sport: running shoes on as soon as you get home.
- Languages: five new words with the first coffee.
- Writing: one sentence of the draft when you open the laptop.
- Home: one thing back in its place when you come through the door.
Choose a moment, not a time window. "In the evening" is not a cue.
Step 3. Keep the action tiny at first
Make the first version so small it is almost silly: one glass, two minutes outside, one page, five minutes of exercise. At the start you are building the repetition, not the result. A tiny action survives a tired Wednesday. A big one often doesn't. You can always do more on a good day. You just don't promise it.
Step 4. Expect months, and plan for them
Decide now that the first two months are for building, not for judging. Put one check-in in your calendar about ten weeks out, and nothing earlier. At the check-in ask a single question: is it taking less effort to start than it did in week one? If yes, you are on your way, whatever the day count says. If not yet, make the step smaller or the cue clearer, and keep going.
Step 5. Don't restart the count after a missed day
In Lally's study, missing a single day did not seriously harm habit formation. So a missed day is a day off, not a reset. Count the days you did it, not the days in a row. The next day, do the tiny version at the usual cue. No paying it back.
That study looked at single missed days. A longer break, such as a trip or the holidays, is a different case, and we wrote a plan for it: how to keep your habits going through the holidays.
Step 6. Make it bigger only when it feels easy
Once the tiny version takes almost no effort to start, add a little: two pages instead of one, ten minutes outside instead of two. If the bigger version feels like a fight, go back to the small one for a while.
Four habits, from easier to harder
These rows go from the habit that asks least of you to the one that asks most. Each has a tiny version to start with and a stable cue. The table gives no number of days, because the studies above give none for any single habit.
| Habit | A tiny version to start with | A stable cue |
|---|---|---|
| Drinking water after breakfast | Fill a glass and drink it | When you put your breakfast plate in the sink |
| A 10-minute walk | Two minutes: shoes on and out of the door | When you get home from work, before you sit down |
| Reading 10 pages | One page | When you get into bed, with the book already on the pillow |
| A 30-minute workout | Workout clothes on and five minutes of any exercise | When the morning alarm goes off, with the clothes laid out the night before |
Common mistakes
- Treating 21 days as a deadline. On day 22 the habit has neither failed nor finished. The averages are about two months, and the range is wide.
- Starting at full size. A 30-minute workout from day one is the version most likely to fall over on a tired Wednesday.
- Using a mood or a time window as the cue. "When I feel like it" and "in the evening" don't give you a moment to act on.
- Restarting the count. One missed day is not the end of the habit.
- Comparing your day count with someone else's. The studies found wide differences between people.
Questions people ask
So is it 21 days or 66 days?
Neither is a deadline. 21 days is a myth. 66 days is an average from one study, and a review of 20 studies found a median of about 59 to 66 days, with individual times up to 335.
What if it has been two months and it still isn't automatic?
Then you are within the range: the review found individual times of up to 335 days. Check whether it is easier to start than in week one. If not, make the step smaller or the cue clearer.
Do I have to do it every day?
The study behind the 66-day average followed a new daily habit, and in it, missing one day did not seriously harm the process. So aim for every day, and treat a miss as a miss, not a reset.
How do I know when it is a habit?
When you do it without talking yourself into it, and you start almost before you have decided. In Lally's study the habit grew more automatic over time until it leveled off, so expect a slope, not a switch.
Let Aimora carry the slow part
A habit that takes months needs something that still works in month three. Aimora gives you one small step a day, built from your goal, the time you have and where you are right now. If a step is still too much, "Too much? Make it smaller" brings it down. When you are steady, the steps grow with you.
If you miss several days, Aimora doesn't ask you to catch up. It starts again from the smallest step. You can follow your streak and progress in the app, and see today's step on an Apple Watch or in a widget. It won't hurry the habit, but it keeps each day's step small enough to show up for.
