Habit Formation Calculator

A new habit usually takes about two months, but anywhere from 18 to 254 days

Researchers at University College London (UCL) asked 96 people to pick one new behavior they wanted to adopt (mostly eating, drinking or exercise), do it every day in the same situation (for example, after breakfast), and rate every day for 12 weeks how automatic it felt (Lally 2010). Among the 39 people whose scores fit a curve well, half reached 95% of their automaticity curve within 66 days, and the middle half took between 39 and 102 days. The fastest person got there on day 18 and the slowest on day 254.

The calculator above turns that result into your own dates. It counts your start date as day 1, shows when day 66 falls and when the middle-half days 39 to 102 fall, and uses a progress bar and a chart to show where today sits on the curve. In the habit calendar, tapping a day checks it off, and your streak and missed days appear alongside. For example, for a habit started on October 1, day 66 is December 5.

Why it helps to know when a habit settles in

People often say that "it takes 21 days to form a habit." On the median curve from the study above, day 21 is only about 65% of the way to the curve's plateau, and reaching 95% takes roughly two more stretches of that length. If you still have to remind yourself after three weeks, you have not failed; the curve is simply still climbing. Writing the finish date on a calendar makes it easier to expect that stretch and stay with it.

• Setting a New Year or new semester resolution: for a habit started on January 1, 2027, day 66 is March 7. If you planned the resolution for "one month," you would stop right in the steepest part of the climb.
• Starting a harder habit such as exercise: participants who chose exercise had the longest median, 91 days. Planning for three months from the start makes a mid-course slowdown less discouraging.
• Wondering around week three why it is still hard: check which day you are on and what percent of the curve that is. It tells you whether the effort you feel is within the normal range.
• Missing a day and losing motivation: leave the missed day in your habit calendar and check today. In the study, a single missed day did not materially affect the process.

How it is calculated

1. Counting days: your start date is day 1. The date of day n = start date + (n − 1) days.
2. When it settles in: the study median for the habit type you pick (other or not sure: day 66, eating: day 65, drinking: day 59, exercise: day 91).
3. 5 or 3 times a week: the study participants did it daily, so the date of repetition n is stretched to day ⌈n × 7 ÷ times per week⌉. The resulting dates and ranges are labeled as estimates.
4. Automaticity curve: the diminishing-returns curve the researchers used, y = a − b·e^−cx (x = day). It takes the median shape of the 39 participants (a = 35, b = 30) and sets c so that the curve reaches 95% on the date above. The curve starts at 14% of the plateau (a).
5. Habit calendar: counts the days you checked, the share of past days you checked, the streak running back from today (or from yesterday if you haven't done it yet today), and your longest streak.

Study results by habit type

These are the 39 well-fitting participants, split by the behavior they chose. The last column is the median day if you started on October 1.

The exercise group took about one and a half times longer than the eating and drinking groups, which the researchers considered consistent with exercise being a more complex behavior. But the differences between types were not statistically significant (p = 0.328), and the researchers noted that the study was not large enough to compare subgroups. So when you pick a type, the calculator shows that type's median as the main date but also states the overall median of day 66 in the same sentence. The fastest (day 18) and slowest (day 254) cases refer to all participants, whatever their habit type.

The phases of the automaticity curve

The curve climbs steeply at first and then flattens out. That means a day of repetition early on adds much more than a day late in the process. The table shows the first day the median curve reaches 50%, 80% and 95% of its plateau, and where day 21 falls.

On the overall median curve, half of the build-up happens in the first 13 days, the next 21 days take it up to 80%, and filling in the last 15% takes another 32 days. The phase bands on the chart are these four stretches. The researchers explain that they used 95% rather than 100% because the curve never actually touches its plateau, so counting to 100% would give an absurdly large number of days, and on the chart 95% is where the curve visibly flattens.

Habits you don't do every day

Every participant in the study acted daily. Hardly any study has measured habits done only a few times a week, so the calculator assumes the same number of repetitions gives a similar degree of automaticity and stretches the days accordingly. The researchers themselves had wanted to look at the link between repetitions and automaticity, but because some days were not recorded they analyzed by calendar day instead.

If you exercise three times a week, the 66th repetition lands about five months out. A 2024 systematic review (Singh 2024) also lists how often the behavior is done among the factors affecting habit strength. In short, doing the same behavior more often, in the same situation, tends to settle it in sooner.

Where the 21-day rule came from

The claim that "it takes 21 days to form a habit" is generally traced to the 1960 book Psycho-Cybernetics by plastic surgeon Maxwell Maltz. He described patients taking about 21 days to get used to a new face after surgery, and amputees feeling a phantom limb for about 21 days, and wrote, "It usually requires a minimum of about 21 days to effect any perceptible change in a mental image." This was not a study of habits, and the original sentence said "a minimum of about" 21 days. In a 2012 post, UCL researchers Gardner and Meisel wrote that it is unclear how that plastic-surgery anecdote came to be generalized to habit formation as a whole.

That does not mean nothing changes by day 21. On the curve above, you are already about two-thirds of the way up by then, so you may well feel you are trying much less hard than at the start. But if you stop there, the remaining third never gets built.

Miss one day and start over?

No. In the Lally 2010 study, 55 participants skipped a day after three days in a row, 140 times in all. Right after the skipped day, their automaticity score dropped by an average of 0.29 points (on a 0 to 42 scale), but when they did the behavior again the next day, it was 0.55 points higher than before the skip. Comparing that with the 0.79-point rise after two days in a row, the researchers concluded that missing one opportunity did not materially affect the habit formation process.

This is why the calculator's habit calendar keeps a missed day on the grid as a hatched box. Even if your streak breaks, your count of checked days and repetitions keeps building. What matters is doing it again the day after you miss.

Conditions that help a habit settle in faster

• Do it in the same situation: habits form as a situation (the cue) and a behavior are paired again and again. Participants in Lally 2010 also fixed a daily situation, such as "after breakfast."
• Attach it to something you already do: in a flossing study (Judah 2013, 50 people), those who flossed after brushing developed a stronger habit than those who flossed before brushing, and eight months later their habit was still stronger and they flossed more often.
• Try mornings: in a study of 48 university students who stretched for 90 days (Fournier 2017), the morning group was estimated to reach automaticity at about day 106 and the evening group at about day 154. These are estimates extrapolated along the curve beyond day 90, and come from a single study, so treat them as a rough guide only.
• Pick a behavior you chose yourself and can enjoy: a systematic review of 20 studies and 2,601 people (Singh 2024) listed frequency, time of day, habit type, whether the habit was self-chosen, and how positively people felt about the behavior (affective judgments), among other factors that affected habit strength. It reported that morning habits and self-chosen habits were generally stronger.
• Start small and simple: complex behaviors may become automatic later, as the exercise group took the longest. Instead of "exercise for an hour," start with a size that has a clear cue and action, such as "walk for 15 minutes before breakfast."

Checking how automatic it feels

Habit researchers measure how settled a habit is not by the number of days but by how automatically you do it. A short four-item scale is the Self-Report Behavioural Automaticity Index (SRBAI, Gardner 2012). After the stem "Behaviour X is something…", you rate your agreement with four statements from 1 to 7.

1. …I do automatically.
2. …I do without having to consciously remember.
3. …I do without thinking.
4. …I start doing before I realize I'm doing it.

Write down the average of the four scores every week or two. At first it rises quickly, and at some point it barely rises at all; if that score is above the midpoint (4), it is close to what the study calls the curve flattening out. The Lally 2010 researchers also treated a plateau below the midpoint (every item answered "neither agree nor disagree") as a habit that had not formed. If your score is still rising after day 66, you are still climbing, and if it flattened above the midpoint before day 66, the habit settled in faster than average.

What other studies found

The Singh 2024 systematic review included 20 studies (2,601 people) on physical activity, drinking water, taking vitamins, flossing, healthy eating, reducing sitting time and more. Four of them reported how long it took for a habit to form: the medians were 59 to 66 days, the means were 106 to 154 days, and individuals ranged from 4 to 335 days. The four studies differ in behavior and in how they calculate, so these numbers cannot be merged into one. That is why the calculator uses the median from Lally 2010, which analyzed daily measurements with a curve, as its main date and shows the range alongside it.

Common mistakes

• Quitting at day 21 because "it's done": on the median curve, day 21 is about 65% of the plateau. Stopping there skips the longest, last stretch.
• Starting the count over after missing a day: keep your start date and simply do it again the next day. Even if the streak breaks, the repetitions you built stay.
• Reading 66 days as a promise: 66 is a median, so half of the participants took longer. Even day 102 is still inside the middle half.
• Doing it "whenever I have time," with no cue: if the situation changes every day, the cue and the behavior never get paired. Fix a time or an earlier action to hang it on.
• Trusting old "difficulty and motivation score" multipliers: stretching the days by a motivation score or by the number of earlier attempts has no basis in the research, so this calculator does not use it.

Limits and cautions

• The 66 days is the result for the 39 best-fitting participants, who were mostly graduate students. There is no guarantee it applies as is to people of other ages, occupations or behaviors.
• Automaticity was measured with a self-rated score (a self-report habit index). The curve on the chart is the participants' median shape stretched to your dates, not a value measured from your own scores.
• The dates for 5 or 3 times a week are an estimate stretched by repetition count and have not been confirmed by research.
• Behaviors that are hard to quit, such as smoking, drinking or drugs, and habits for managing a medical condition are outside this calculator's scope. Please get help from a health professional or a specialist service.
• Everything is calculated inside your browser. If you turn on saving, it stays only in this browser on this device. A link made with "Copy link" carries the name, start date and checks after the # in the address, which is not sent to a server, and the page removes it from the address as soon as it opens.

As of Lally 2010 was checked against the full-text PDF on 2026-10-14, Gardner 2012 against the Europe PMC full text the same day, and Singh 2024, Fournier 2017 and Judah 2013 against their Europe PMC abstracts the same day.

Sources

• 01Lally P, van Jaarsveld CHM, Potts HWW, Wardle J. How are habits formed: Modelling habit formation in the real world. Eur J Soc Psychol 2010;40(6):998–100982 of 96 participants analyzed, 39 with a good model fit. Median 66 days to 95% (Q1 to Q3: 39 to 102; range 18 to 254), eating 65, drinking 59, exercise 91 days (p = 0.328), 0.29-point drop right after a missed day
• 02Singh 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):248820 studies, 2,601 people. Medians 59 to 66 days, means 106 to 154 days, individuals 4 to 335 days. Factors include frequency, time of day, type, self-chosen habits and affective judgments
• 03Gardner B, Abraham C, Lally P, de Bruijn GJ. Towards parsimony in habit measurement: testing the convergent and predictive validity of an automaticity subscale of the Self-Report Habit Index. Int J Behav Nutr Phys Act 2012;9:102Four automaticity items (SRBAI), answered on a 1 to 7 scale
• 04Fournier M, et al. Effects of circadian cortisol on the development of a health habit. Health Psychol 2017;36(11):1059–106448 university students, 90 days of stretching. Extrapolated time to automaticity: 105.95 days in the morning, 154.01 days in the evening
• 05Judah G, Gardner B, Aunger R. Forming a flossing habit: an exploratory study of the psychological determinants of habit formation. Br J Health Psychol 2013;18(2):338–35350 people. Flossing after brushing gave a stronger habit than flossing before, and was still stronger and more frequent eight months later
• 06UCL Health Behaviour Research Centre blog: Busting the 21 days habit formation myth (2012)Quotes the preface of Maltz 1960 Psycho-Cybernetics: "a minimum of about 21 days"

Frequently asked questions

Tools to use with this one