Lally et al. (2010): How Are Habits Formed

Lally and colleagues had volunteers pick a new daily eating, drinking or exercise behaviour tied to a fixed cue and tracked how automatic it felt over 12 weeks. Automaticity grew along a curve that plateaued after a median of 66 days, with a huge range of 18 to 254 days depending on the person and behaviour. Missing a single day made little difference to the curve, which undercuts the idea that one slip resets a habit.

CitationLally, 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.
Study typeLongitudinal real-world study with daily self-report over 12 weeks
Key findingMedian 66 days to peak automaticity, range 18 to 254 days
Slip findingMissing one opportunity did not materially derail habit formation
Sample96 volunteers; 82 provided enough data for curve modelling

What the study found

Each participant chose one behaviour to repeat daily in the same context, drinking water with lunch, a 15-minute walk after dinner, and rated daily how automatic it felt. Fitting curves to the data showed automaticity rising steeply at first and then flattening, reaching its plateau at a median of 66 days. Simpler behaviours became automatic faster than effortful ones like exercise, the individual range was enormous, and occasional missed days did not visibly dent the curve.

Why it matters

For procrastination, habits are the endgame: a behaviour that runs on cue does not need the motivational battle that avoidance wins. This study sets honest expectations, weeks to months rather than the mythical 21 days, and delivers two useful rules: anchor the new behaviour to a fixed daily cue, and treat a missed day as noise rather than failure. That second finding is the empirical cousin of the self-forgiveness result, never miss twice beats never miss.

Caveats

The sample was 96 self-selected volunteers, mostly students, tracking one self-chosen behaviour each, and automaticity was self-rated rather than observed. Dropout was substantial and the curve-fitting excluded poor fits, so the 66-day median describes the modelled subset. The 18 to 254 day range is the honest headline; quoting 66 days as a rule overstates the precision.

Source

Peer-reviewed study in the European Journal of Social Psychology, the primary empirical source for modern habit formation timelines.

https://doi.org/10.1002/ejsp.674

Related

Habit StackingGollwitzer & Sheeran (2006) on if-then plans

Frequently asked questions

How long does it really take to form a habit?

In this study, a median of 66 days to reach peak automaticity, but the range ran from 18 to 254 days depending on the person and the behaviour. Easy actions habituate in weeks; effortful ones like exercise can take many months.

Does missing a day ruin a habit?

No. In the data, missing a single opportunity had no meaningful effect on the habit curve. The practical rule the finding supports is simple: one miss is noise, so restart the next day rather than declaring the streak dead.

Stuck right now?

Three minutes: find out what kind of procrastinator you are and get a plan for this week.

Take the quiz

One practical email a week

A technique, a study worth knowing, and one thing to try. No streaks, no guilt, unsubscribe any time.