IVYH Journal
How long does it take to form a habit? Forget the countdown
A habit has no universal finish line. Why 66 days isn't a new rule, and an unbroken streak isn't the only sign of change.

In this guide
Day 22 is an odd deadline. If your new habit still needs a reminder, have you done something wrong? When you've been told that three weeks is enough to make a habit stick, it's easy to question your discipline rather than the promise.
There is no universal 21-day threshold for building a habit. Nor is there a 66-day one, the figure often offered in its place. Research describes a gradual, highly variable process: with repetition, an action can begin to feel more automatic. The calendar alone cannot tell you how far you've come.1, 2
When an action starts to come naturally
Doing something often and doing it out of habit are closely related, but they aren't quite the same. You might walk every lunchtime and still have to talk yourself into it each day. As a behaviour becomes more automatic, getting started needs less conscious prompting.
In a widely cited study by Phillippa Lally and colleagues, 96 people were asked to choose a new eating, drinking or exercise behaviour and try to repeat it in the same everyday context for twelve weeks. They reported both whether they had done it and how automatic it felt. The researchers were tracking self-reported automaticity, not whether a morning routine improved anyone's health.1
The distinction matters. A tick on the calendar records something you did. It doesn't necessarily show how readily the action will come to you next time.
What those 66 days actually meant
The researchers fitted mathematical curves to participants' reports. Among the 39 people whose data fitted the model well, the median modelled time to reach 95 per cent of each individual's estimated automaticity plateau was 66 days. The modelled range was 18–254 days.1
That's a more specific finding than “a habit takes 66 days”. It applied to a selected part of the group and a particular definition used in the analysis. The study also lasted 84 days: the longer estimates came from the model, not from following people for 254 days.
A 2024 systematic review included 20 studies and 2,601 participants, but only four studies directly reported how long habit formation took. Methods and measures differed too much for the researchers to pool the time estimates in a meta-analysis. Many of the studies also had a high risk of bias.2
Giving all those participants the same finish line would be misleading. A walk, a glass of water and a workout aren't identical behaviours. Nor do two people make the same everyday choices in identical circumstances.
Give the habit a place, not just a start date
A more useful focus is the situation that will remind you. “I'll read more” describes an intention. “After I've cleared breakfast away, I'll open my book” also identifies an occasion. The second version makes it clearer where the action belongs.
In Lally's study, the behaviour was linked to a recurring situation. Repetition in a consistent context is central to the study's account of habit formation. That doesn't mean one particular cue works best for everyone, or that an existing routine always beats a time of day.1
You can try the idea without turning it into an ambitious life overhaul: choose something you genuinely want to return to and an occasion that's already there. A glass of water as you set out lunch, or a little reading after breakfast. These are planning examples, not promises that a particular habit will be complete by a particular day.
Missing once doesn't mean going back to zero
An unbroken streak is easy to count. Once it's broken, it can feel as though the work disappeared with it.
In Lally's study, missing a single opportunity did not appear to materially disrupt habit formation. The analysis looked at specific single omissions following three consecutive occasions when the behaviour had been performed. It doesn't show that longer breaks are irrelevant, but it offers no support for treating every missed occasion as a complete reset.1
That's a more useful thought to keep than a perfect streak. The next opportunity is still there, even when yesterday's tick isn't.
A ritual can be yours before it feels familiar
A ritual can be something you choose with care: a drink, a pause, a small transition between work and the rest of the day. You don't need to do it automatically to enjoy it. In these studies, the researchers were asking a different question: how automatic the behaviour felt.
At IVYH, daily rituals are worth exploring beyond what's in the glass. In IVYH Journal, we look at habits, ingredients and considered everyday choices.
The next time you want to start something new, try changing the question. Not 'what date should this be a habit by?', but 'which part of my ordinary day do I want to return to?'. The second question gives you something to work with before day one.
Sources
- Lally P. et al. How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology. Published online 2009; journal issue 2010.
- Singh B. et al. Time to Form a Habit: A Systematic Review and Meta-Analysis of Health Behaviour Habit Formation and Its Determinants. Healthcare, 2024;12(23):2488.
