Short answer: far longer than 21 days, and it depends enormously on the habit. The most-cited research found a median of around 66 days — with some people settling into a new behaviour in under three weeks and others taking the better part of a year. Here's where the famous number came from, what was actually measured, and why the answer is more forgiving than it sounds.
Where the 21-day rule came from
It came from a plastic surgeon. In 1960, Maxwell Maltz published Psycho-Cybernetics, in which he noted that patients seemed to need "a minimum of about 21 days" to get used to seeing a new face in the mirror, or to stop feeling a phantom limb.
That's it. It was an observation about adjusting to a changed body — not a study of habit formation, not an experiment, and not a claim that three weeks makes a behaviour stick. The number survived because it's short, tidy and hopeful. None of those make it true.
What the research actually found
The study people usually mean was run at University College London and published in 2010. Participants chose one simple daily behaviour — a piece of fruit with lunch, a short run, a glass of water — and reported each day on how automatic it felt.
Three findings are worth carrying with you:
- The median was about 66 days. Not 21. Roughly two months before the behaviour felt like something they simply did, rather than something they decided to do.
- The range was enormous. From around 18 days to over 250. The averages hide how differently this goes for different people and different habits.
- Difficulty mattered more than willpower. Drinking water after breakfast became automatic quickly. Doing fifty sit-ups before dinner took far longer. The gap was in the behaviour, not the person.
It's one study, with fewer than a hundred participants, all choosing their own habits — so hold the number loosely. But its direction is clear and it is much better evidence than a 1960 aside about post-surgical adjustment.
The finding that actually changes things
Buried in that research is the part worth pinning above your desk: missing a single day had no measurable effect on the eventual outcome.
One missed day did not reset progress. It did not undo the weeks before it. The curve carried on more or less where it left off — as long as the person came back.
This matters because the streak is what breaks most people. Not the missed Tuesday — the story told about the missed Tuesday. Well, I've ruined it now. The evidence says you haven't. You've had a Tuesday.
What to do with a two-month answer
If automaticity takes months rather than weeks, then the strategy that works is the one you can still be doing in month two.
- Make the first version embarrassingly small. A habit that takes two minutes survives a bad week. One that takes forty does not. You can grow it later; you cannot grow one you've abandoned.
- Attach it to something that already happens. After I pour my coffee, I write one line. The existing routine does the remembering for you, which is most of the work in the early weeks.
- Pick same-time over same-amount. Consistency of the cue matters more than the size of the effort. Ten minutes every morning beats an hour whenever you can manage it.
- Plan the miss before it happens. Decide now what a bad day looks like — the two-minute version, not zero. Having a floor means never having to start over.
- Count returns, not streaks. A streak punishes you once and then keeps punishing. Coming back is the skill that actually builds the habit.
Why "just be disciplined" doesn't work
Sixty-six days is a long time to run on willpower, and willpower is exactly the resource that a stressful week drains first. That's why the habits that survive are the ones designed to need very little of it: small, cued by something that already happens, and forgiving of an off day.
It's also why beating yourself up backfires. Shame is a poor motivator over two months — it makes the habit something you're avoiding rather than something you're building. Curiosity holds up better. What got in the way today? is a more useful question than why am I like this?
Inside the app, Coach Roger's daily check-ins and gentle streak tracking are built around exactly this: noticing the pattern without turning a missed day into a verdict.
The bottom line
Forget 21 days. Expect something closer to two months, expect it to vary wildly depending on what you've chosen, and expect to miss days. The research is unusually kind on that last point: one missed day changes nothing measurable. The habit is built by what you do the day after.
Related: does journaling actually help — one of the smallest habits worth trying, and one of the best studied.
This guide is general wellness information, not medical advice. If you're struggling with motivation in a way that feels heavier than habit-building, please talk with a healthcare professional.