It depends — a lot more than the headlines suggest. In the most widely cited study on the subject, it took people anywhere from 18 to 254 days for a new habit to become automatic. The famous 21 days didn’t come from any study at all.
And before you close the tab thinking ‘254 days, how awful’: the good part is coming up, and it’s that skipping a day doesn’t ruin it.
The 21 days came out of an operating theatre
The number has a rather more curious origin than you’d think. In 1960, a plastic surgeon called Maxwell Maltz published a self-help book in which he described how his patients took about 21 days to get used to their new face after surgery. That’s how Scientific American tells it, adding the key detail: nobody ran any experiment to check it.
The book then applied those three weeks to almost everything — getting used to a new house, changing your mind — and over the years the phrase was polished down to ‘it takes 21 days to form a habit’. Which wasn’t about habits, and which nobody had measured.
It’s not a bad number, mind you. It’s round, it fits in a headline and it sounds like a reasonable effort: not three days, which nobody believes, nor a year, which is daunting. Its only problem is that it isn’t true.
What was actually measured
It took almost half a century for someone to measure it properly. A team at University College London followed 96 volunteers for twelve weeks. Each of them chose something small to do every day, always at the same moment — a piece of fruit with lunch, a bottle of water, a short run before dinner — and every day they recorded how much effort it took.
The result, published in the European Journal of Social Psychology, was a huge range: from 18 to 254 days until the thing became automatic. The number that later became famous is 66, but latching on to the round figure is repeating the mistake of the 21 with a different number.
The interesting part is why the difference is so big. Phillippa Lally, the lead author, puts it with plenty of common sense: learning something completely new takes longer than remembering to drink more water in the morning. The simpler the habit, the sooner it stops being an effort — which won’t surprise anyone who has tried both.
Skipping a day doesn’t ruin it
This is the part of the study we like best, and the part that gets mentioned least.
Did you know? The authors put it in writing, in the abstract of the paper itself: ‘Missing one opportunity to perform the behaviour did not materially affect the habit formation process.’ One stray day doesn’t send you back to square one.
What does do damage is what we tend to do after that day: decide it doesn’t matter any more and give up altogether. That has a name, and we cover it in what to do when you break a streak. The problem was never the Tuesday you missed. It’s the ‘well, I’ve ruined it now’ on Wednesday.
It’s exactly the idea behind soft wellness: what counts is what you do most of the time, not what you do all of the time.
What the study doesn’t say
Here comes the part the headlines tend to skip, because when the science comes with caveats, we say so.
It’s what each person reported about themselves. Nobody measured anything from the outside: each volunteer rated how much effort it took. That’s a reasonable way to measure something like ‘I do it without thinking’, but it’s subjective.
Not everyone gave usable data. Of the 96 who started, the model only fitted well for 39. The conclusions come from that group.
And the study lasted 84 days. So nobody actually saw the 254: it’s what the model calculates that person would have taken. A projection, not an observation.
None of this invalidates it. It just means reading it for what it is: the best data there is, not a law.
So how do I know I’ve got it?
Not by the calendar. The sign is the one we gave in the stacking trick: the day you do it without remembering you were trying to.
And to get there sooner, what worked in the study is the same as what works outside it. Do it at the same moment every time, attached to something you already do. Keep it in sight — Lally gives the example of keeping fruit within reach, which is the see-it-eat-it trick in other words. And if it stalls, don’t push harder: make it smaller. If you’re missing days, instead of a long run every day, a short jog three times a week. She says that too.
Start today: pick the habit you’ve been trying to build and forget about counting days. Attach it to something you already do, and on the day you forget, just carry on the next.
Frequently asked questions
How long does it take to form a habit?
There’s no number that applies to everyone. In the most widely cited study, by a team at University College London, it took participants between 18 and 254 days for a new habit to become automatic. The simpler the habit, the sooner it stops being an effort.
Where does the 21 days idea come from?
From a 1960 self-help book. Its author, the plastic surgeon Maxwell Maltz, wrote that his patients took about 21 days to get used to their new face after surgery. He wasn’t talking about habits, and nobody ran an experiment to check it.
What happens if I skip a day?
Nothing serious. The study’s authors wrote it in the abstract: missing one opportunity to perform the behaviour did not materially affect the habit formation process. What does do damage is giving up after that day.
Why are some habits harder than others?
Because remembering to drink water isn’t the same as going for a run. Simple habits you can repeat often become automatic sooner; those that take effort or involve learning something new take quite a lot longer.
How do I know a habit has formed?
Not by the calendar: the sign is the day you do it without remembering you were trying to. To get there sooner, it helps to do it at the same moment every time, attached to something you already do, and to make it smaller if it stalls.