The surgeon who counted days
If you've ever started running or keeping a journal, you've probably been told: stick with it for 3 weeks, and then it will happen on its own. 3 weeks is 21 days, and the number sounds like someone calculated it on purpose.
Someone did calculate it, but not for that. In 1960, American plastic surgeon Maxwell Maltz published the book "Psycho-Cybernetics." In it, he shared an observation from his practice: a patient needs "a minimum of about 21 days" to get used to a new face in the mirror. He gave roughly the same time frame for the sensation of a phantom limb after amputation.
Maltz wrote about adapting to a changed body, not about how a new action becomes established. And he wrote "a minimum of about." The book was retold for decades - in trainings, articles, and seminars. Along the way, the caveat fell away, the topic changed, and all that remained was: "21 days - and your habit is ready."
Any habit forms in 21 days - it's proven.
The number comes from a surgeon's 1960 book about getting used to your appearance after surgery. It was a doctor's observation, not a habit experiment.
What happened when habits were measured
In a 2010 University College London study, 96 people chose one simple action - a glass of water with lunch, a walk after dinner - and repeated it every day in the same situation. Every evening, they noted how automatic the action felt: without thinking, just happening.
The curve looked similar for most people: rapid growth at first, then a flatter and flatter rise before reaching a plateau. The median was 66 days. But more interesting than the median was the range: the fastest participant reached a plateau on day 18, while the model estimated that the slowest needed about 254 days.
The type of action mattered too: drinking a glass of water became established faster than an exercise like squats. And missing 1 day didn't noticeably disrupt the curve - contrary to the rule "don't break the chain."
So "how many days does it take?" has no single answer. Your morning stretch might become automatic in a month, while putting your phone away before bed could take 6 months.
66 days is a benchmark from one small study, where participants rated their own automaticity. The direction is reliable: weeks and months, not 3 weeks. There is no exact number.
So 21 days isn't dangerous in itself, but as a deadline. On day 22, it's easy to decide something is wrong with you - when you're simply in the middle of a normal range.
Why doesn't a 3-week deadline work for a habit?
What helps a habit stick without extra pressure?
Sources
- Maxwell Maltz, "Psycho-Cybernetics," 1960. Lally et al., "How are habits formed: Modelling habit formation in the real world," European Journal of Social Psychology, 2010
Next in the “Habits” series
3 articles in full on the site, 4 more in the app
