How Long Does It Take to Build a Habit? The Honest Answer
Updated 26 August 2026 · 8 min read
Research from University College London (Lally et al., 2009) found new habits took between 18 and 254 days to become automatic, with a median of around 66 days. The famous "21 days" figure comes from a 1960 self-help book, not a study — but three weeks is still a genuinely useful window for building structure and momentum.
Key takeaways
- Lally et al. (2009, UCL) found new habits took 18 to 254 days to become automatic, with a median of around 66 days.
- The popular “21 days to form a habit” line is a misreading of Maxwell Maltz’s 1960 book Psycho-Cybernetics, which described self-image after surgery, not habit studies.
- Missing a single day made no meaningful difference in the Lally data — returning the next day matters more than a perfect streak.
- Simpler behaviours became automatic faster; same cue, same time, and low friction beat motivation.
- Twenty-one days is still a useful starting window for structure and proof of fit, not a finish line for automaticity.
Ask the internet how long it takes to build a habit and you'll get the same confident answer everywhere: 21 days. It's on posters, in podcasts, in a thousand app onboarding screens. It's also not what the research says. This article gives you the honest version — where the 21-day figure actually came from, what the best available study found, and why a 21-day window is still worth using, as long as you're clear about what it's for.
Where the 21-day myth came from
The “21 days to build a habit” claim comes from Maxwell Maltz’s 1960 book Psycho-Cybernetics, not from a habit study. Maltz, a plastic surgeon, observed that patients often took about three weeks to adjust to a change in appearance. He described a minimum of about 21 days for self-image to catch up with reality — not a rule for forming any behaviour.
It requires a minimum of about 21 days for an old mental image to dissolve and a new one to jell. — Maxwell Maltz, Psycho-Cybernetics (1960)
Note two things. First, this was a clinical observation about self-image, not a controlled study of habit formation. Second, Maltz said a minimum of about 21 days. As the book was retold and simplified over decades, "a minimum of about 21 days to adjust your self-image" became "21 days to build any habit" — a much stronger claim that Maltz never made and no study has ever supported.
What the research actually says: 18 to 254 days
The best-known real-world study is Lally et al. (2009/2010) in the European Journal of Social Psychology. University College London researchers tracked volunteers repeating a chosen daily behaviour (for example a glass of water with lunch). Time to automaticity ranged from 18 to 254 days; the median was around 66 days; simpler behaviours formed faster; missing one day made no meaningful difference.
- **Time to automaticity ranged from 18 to 254 days.** Some habits locked in within three weeks; others were still building after eight months.
- **The median was around 66 days** — roughly two months, not three weeks.
- **Missing a single day made no meaningful difference.** One skipped day did not derail habit formation. Consistency over the whole period mattered far more than a perfect streak.
- **Simpler behaviours became automatic faster.** Drinking a glass of water habituated more quickly than doing exercise, which is exactly what you'd expect: the smaller the action, the sooner it stops requiring willpower.
So the honest answer to "how long does it take to build a habit?" is: it depends on the habit, and probably longer than 21 days — but a single missed day won't undo your progress, and the easier you make the behaviour, the faster it sticks.
Why 21 days is still a genuinely useful starting window
Twenty-one days is still useful because it is long enough to install structure, ride the steepest early automaticity gains, and test whether a routine survives a real Tuesday — it is not long enough to finish most habits. Treat three weeks as a starting window. The median in the Lally data is still around 66 days.
- **Structure.** Three weeks is long enough to run a real routine — the same meals planned, the same shopping list, the same workout days — rather than a one-week novelty.
- **Momentum.** In the Lally data, automaticity gains were steepest in the early weeks. The first 21 days are where each repetition does the most work.
- **Proof of fit.** Three weeks tells you whether a routine survives contact with your actual life — the busy Tuesday, the weekend away, the evening you can't be bothered. A plan that survives 21 real days is a plan worth keeping.
This is exactly why My21Reset is called a reset, not a transformation. Twenty-one days won't make a habit automatic — the research is clear on that — but it will get the hardest part done: the setup, the early repetitions, and the evidence that the routine fits your life. Day 22 onwards is repetition, and repetition is the easy part once the structure exists.
What makes habits form faster
Habits form faster when the behaviour is tied to the same cue at the same time every day, friction is low, the action starts easier than feels necessary, and missed days are treated as irrelevant. Motivation is not the main lever in the Lally data — design is.
| Lever | Why it works | Example |
|---|---|---|
| Same cue, same time, every day | Habits are cue-response links. A stable cue (after breakfast, on arriving home) gives the behaviour something to attach to. | Prep lunch straight after the kettle goes on each morning |
| Lower the friction | The fewer decisions and steps between cue and action, the more often it happens. | A [prepped week of meals](/guides/meal-prep-for-the-week) beats deciding what to cook at 7pm |
| Start easier than feels necessary | Simple behaviours became automatic fastest in the Lally study. You can always scale up later. | Two short [home workouts](/guides/beginner-home-workout-plan) a week, not six gym sessions |
| Don't chase a perfect streak | Missing one day made no meaningful difference in the data. Restarting quickly matters; guilt doesn't. | Miss Tuesday? Do Wednesday as planned. Nothing is broken. |
Notice what isn't on the list: motivation. The people who succeeded in the research weren't the most motivated — they were the ones whose behaviour was tied to a reliable daily cue and easy enough to repeat. Design beats discipline.
What to do after day 21
After day 21, keep the same cue and time, repeat before you add new behaviours, rotate a small set of meals so boredom does not kill the routine, and expect weeks four to six to feel flatter than the first three. If the median is around 66 days, day 22 is roughly a third of the way there.
- 1.**Keep the cue, keep the time.** Whatever anchor carried you through the first three weeks, don't renegotiate it. Changing the cue restarts the clock.
- 2.**Repeat, don't expand.** The temptation after a good three weeks is to add more. Resist it until the current routine feels automatic. Depth before breadth.
- 3.**Keep a small rotation.** Boredom kills routines faster than difficulty. A handful of easy go-to recipes you rotate is more durable than a different meal every night.
- 4.**Expect the dip.** Weeks four to six often feel flatter than weeks one to three — novelty has worn off and automaticity hasn't fully arrived. This is normal, and it's where the one-missed-day finding matters most: keep showing up imperfectly.
The honest summary
The 21-day rule is a misreading of a 1960 book about self-image. The real research says habits take anywhere from 18 to 254 days to become automatic — around 66 in the middle — and that a missed day here and there doesn't matter. But three weeks remains the right size for a first commitment: big enough to build real structure and momentum, small enough that you'll actually finish it. Use 21 days to start the habit and prove the routine fits. Then just keep going — which is exactly what a well-designed 21-day vegetarian reset is built to set up.
Frequently asked questions
Does it really take 21 days to build a habit?
No. The 21-day figure comes from Maxwell Maltz's 1960 book Psycho-Cybernetics, where it described how long patients took to adjust their self-image — it was never a study of habits. Research by Lally et al. (2009) at UCL found habits took 18 to 254 days to become automatic, with a median of around 66 days.
What did the Lally et al. habit study actually find?
Lally et al. (2009), published in the European Journal of Social Psychology, tracked volunteers repeating a chosen daily behaviour. Automaticity took between 18 and 254 days to plateau, the median was around 66 days, simpler behaviours formed faster, and missing a single day made no meaningful difference to the outcome.
Does missing a day ruin a habit?
No. In the UCL research by Lally and colleagues, missing one day made no meaningful difference to habit formation. What matters is returning to the routine the next day. Perfect streaks are motivating for some people, but they are not what the evidence says habit formation depends on.
Is a 21-day challenge still worth doing?
Yes, if you treat it as a starting window rather than a finish line. Three weeks is long enough to set up structure, build early momentum — the steepest automaticity gains happen in the first weeks — and prove a routine fits your real life. The habit finishes forming in the weeks that follow.
How can I make a habit form faster?
Tie the behaviour to the same cue at the same time every day, make it as low-friction as possible (decide once, prepare ahead), and start easier than feels necessary — simple behaviours became automatic fastest in the research. Then don't panic over missed days; consistency over weeks beats perfection.
What should I do after a 21-day reset ends?
Keep the same cues and times rather than redesigning the routine, repeat before you expand, and rotate a small set of go-to meals and sessions to keep boredom away. Expect weeks four to six to feel flatter than the first three — that's normal, and pushing through it is where habits actually settle in.
Is the 21-day habit rule scientifically proven?
No. There is no study showing that 21 days is the time required to form a habit. The number is a popular simplification of Maltz (1960). The UCL research that actually measured automaticity found a range of 18–254 days and a median of about 66.
Sources
- Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009.
- Maltz, M. (1960). Psycho-Cybernetics. Prentice-Hall.