How Long Does It Take to Form a Habit? What the Research Says (It Isn't 21 Days)

The most cited study on the subject, by Phillippa Lally and colleagues (University College London, 2009), found that a behavior took an average of 66 days to become automatic, but with huge variation: from 18 to 254 days, depending on the person and the habit. The famous "21 days" doesn't come from research on habits. And there's a finding almost nobody mentions: missing a day did not meaningfully disrupt the process.

Where does the 21-day idea come from?

The number is usually credited to plastic surgeon Maxwell Maltz, who wrote in Psycho-Cybernetics (1960) that patients took at least about 21 days to get used to changes, like a new face. It was a clinical observation about adjustment, not a study of habit formation. Over time, the "at least" disappeared and it became a rule.

The problem with the myth isn't just that it's imprecise. It's that many people reach day 22, notice it still isn't automatic, and conclude they failed, when in fact they're partway there.

What did Lally's study find?

Researcher Phillippa Lally and her team followed 96 volunteers for 12 weeks. Each chose a new behavior to do every day, always in the same context: drinking a glass of water after breakfast, eating a piece of fruit at lunch, running for 15 minutes before dinner, and so on.

Every day, participants rated how automatic it felt. The main results:

• An average of 66 days until automaticity leveled off.
• A range of 18 to 254 days (the highest numbers are estimates projected by the statistical model).
• Simple habits became automatic faster. Drinking water was much quicker than doing 50 sit-ups before breakfast.
• Missing one opportunity did not meaningfully affect the habit-formation process.

Lally herself, in later interviews, stressed that the main message isn't the number 66 but that the time varies a lot.

What makes a habit take more or less time?

Size and difficulty

The simpler, the faster. Drinking a glass of water is different from running 5 km. That's why starting with the minimum version helps: you automate the simple part first (see Tiny Habits: Why Starting Small Works (and 20 Examples)).

A stable context

Doing it after the same cue every time (brushing your teeth, your first coffee) speeds things up. Doing it at a different hour and place each day slows it down.

Immediate reward

Habits with some immediate pleasure stick more easily. When the action has no natural reward, creating one, like celebrating or checking it off, can help.

You and your phase of life

Routine, sleep, stress, health, work. A chaotic stretch slows any habit. That isn't a personal failing.

So how do you know the habit has "stuck"?

A few signs:

• You do it without having to remember or decide.
• Skipping it feels strange, like leaving the house without brushing your teeth.
• Doing it takes less effort than at the beginning.

If you haven't gotten there yet, that's not a problem. Keep repeating it in the same context.

I missed a day. Did I lose everything?

No. This may be the most useful finding of Lally's study: missing one opportunity did not meaningfully disrupt habit formation. What does disrupt it is what comes after the miss: the thought "I've blown it, so I give up," which turns one day into a week and a week into abandonment.

If you use streak counters, it's worth thinking about how they handle missed days; we talk about that in Habit Streaks: When They Help, When They Hurt. And if you want a step-by-step way back, see How to Restart a Habit Without Guilt or Making Up for It.

What does it look like in a real day?

Think of a 35-year-old saleswoman we'll call Nina (a made-up example). She decided to stretch for 5 minutes every morning and read that habits take 21 days. On day 21, she still needed an alarm to remember and still found it boring. She concluded it "wasn't for her" and stopped.

Months later, she tries again, knowing the research:

• She shrinks the habit to one neck stretch, after brushing her teeth.
• She doesn't count days to "get there"; she just checks it off when she does it.
• On two days of the week, she forgets. She returns the next day, no drama.
• Around week ten, she notices the stretch has become part of "brushing my teeth."

She didn't do anything extraordinary. She just adjusted her expectations.

Habits and ADHD: does it take longer?

There's no specific number for people with ADHD. Many report that the novelty of a habit fades fast and that keeping up the repetition is the challenge. Visible cues, minimum versions, and reminders that don't nag help. If you suspect you have ADHD or take medication, talk to a health professional.

Frequently asked questions

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

There's no evidence for it. Lally's study (2009) found an average of 66 days, ranging from 18 to 254 days depending on the person and the habit.

How long does it take to form an exercise habit?

It varies a lot. In Lally's study, more demanding habits, such as exercise, tended to take longer than simple ones, like drinking water. Starting with a minimum version helps.

If I miss a day, do I have to start over from zero?

No. In Lally's study, missing one opportunity did not meaningfully affect the process. What matters is returning the next day.

How can I speed up habit formation?

Keep the habit simple, do it in the same context (after a fixed cue), and add some immediate reward, like celebrating or checking it off.

A better goal than a number

Instead of "do it for 21 days" or "for 66 days," try "do the minimum version after the same cue, and come back whenever I miss." The time comes along with it.