01The myth
Where “21 days to form a habit” came from
The 21-day number traces back to Maxwell Maltz, a plastic surgeon who wrote the 1960 bestseller Psycho-Cybernetics. He noticed patients took “a minimum of about 21 days” to get used to a change, like a new face in the mirror.
It was a clinical observation, not a study of habits. Over decades of retelling, “a minimum of about” disappeared and the number became a rule. The 21/90 rule (21 days to build a habit, 90 to make it a lifestyle) grew out of the same idea. Three weeks is a fine length for a 21-day challenge, just not a reason to expect a habit to be locked in by day 22.
02The research
What the research says: 66 days, with a huge range
In 2010, Phillippa Lally and colleagues at University College London asked 96 volunteers to repeat one eating, drinking or exercise behavior daily for 12 weeks and rate how automatic it felt.
Median to automaticity
66days
Lally et al., 2010. Half took longer.
Range between people
18-254
Days. Longer estimates are projections past the 12-week study.
Across 20 studies
59-66days
Singh et al., 2024 systematic review. Range about 4 to 335 days.
| Claim | What it really means |
|---|---|
| 21 days | A loose clinical observation from 1960, not a habit study |
| 66 days | The median in one well-known study; half of people took longer |
| 18 to 254 days | The real spread between people in that study |
| About 2 months | A reasonable expectation for a simple daily habit |
The takeaway: habit formation is a curve, not a finish line. Each repetition in the same context makes the next one a little easier. Simple behaviors, like a glass of water after breakfast, got there faster than 50 sit-ups.
03What real habits show
Most habits never reach day 66. They end in the first two weeks
Lab studies follow volunteers who stick with it. Real life is different. Across 4,404 habits in Daily Pact, 86% of the habits that stopped got their last check-in within two weeks, and the typical solo habit that stopped ended on day 1.
Each week only counts habits old enough to have reached it. n = 1,210 habits with 1-2 friends, 1,579 in groups of 4+, 1,615 alone. Source: The Daily Pact Habits Report 2026.
See the numbers+
| Week | With 1-2 friends | Group of 4+ | Alone |
|---|---|---|---|
| Week 1 | 88.5% | 89.4% | 42.7% |
| Week 2 | 50.3% | 45.2% | 11.8% |
| Week 3 | 26% | 14.8% | 5% |
| Week 4 | 17.5% | 4.3% | 3.1% |
| Week 5 | 14.4% | 3.7% | 3% |
| Week 6 | 11.8% | 1.4% | 2.7% |
| Week 7 | 7.7% | 1.4% | 2.6% |
| Week 8 | 7.5% | 0.9% | 2.1% |
Alone: day 1
With others: day 6
Median day after starting. n = 2,555 habits with no check-in in the last 21 days (516 solo, 2,039 shared). Source: The Daily Pact Habits Report 2026.
See the numbers+
| Alone | With others | |
|---|---|---|
| Median day of last check-in | day 1 | day 6 |
| Stopped on day 0 | 44.6% | 11.4% |
| Stopped in week 1 | 72.1% | 50.2% |
| Stopped by day 14 | 89.1% | 85.2% |
| Stopped by day 30 | 95.9% | 95.9% |
So the honest answer to “how many days to form a habit” has two parts: about two months for the habit to become automatic, and about two weeks to survive the stretch where most habits die. The second one is where your plan should go.
04Streaks
How many people actually reach 21, 30 and 66 days?
Very few, which is why the 21-day myth feels true: people who make it to three weeks are already the exception. 26% of daily habits shared with 1-2 friends reached a 7-day streak, 2.8% reached 30 days, and across all daily habits only 0.2% reached 66 days in a row.
3 days
7 days
14 days
30 days
Daily habits with at least one check-in: 883 with 1-2 friends, 651 alone. 0.2% of all daily habits reached 66 days. Source: The Daily Pact Habits Report 2026.
See the numbers+
| Streak | With 1-2 friends | Alone |
|---|---|---|
| 3 days | 56.9% | 19.5% |
| 7 days | 26.2% | 6.6% |
| 14 days | 9.3% | 0.8% |
| 30 days | 2.8% | 0.2% |
Reached a 7-day streak
1 in 4 habits with 1-2 friends (26%)
1 in 15 habits done alone (6.6%)
A streak is not the habit
A habit can form with gaps. Lally found a single missed day didn't matter. Count days done, not only days in a row, and don't restart from zero after one miss.
05Missed days
Missing one day doesn't reset it. Missing two usually ends it
After a single missed day, 28% of shared habits got a check-in the very next day and about half came back within a week. After two missed days in a row, only 15% came back the next day.
With others
Back next day
28%
Within a week
53%
After 2 misses
15%
Alone
Back next day
19%
Within a week
40%
After 2 misses
10%
n = 4,339 missed days in shared daily habits, 975 in solo ones. Source: The Daily Pact Habits Report 2026.
See the numbers+
| With others | Alone | |
|---|---|---|
| Back the next day after 1 missed day | 28% | 19.1% |
| Back within a week after 1 missed day | 52.7% | 39.5% |
| Back the next day after 2 missed days | 15.1% | 9.6% |
Decide in advance that a miss costs you the day and nothing else. More in never miss twice.
06What speeds it up
Six things that make a habit form faster
Keep it small.
Simple behaviors automate sooner. “5 push-ups” beats “a full workout”. Grow it once it runs on its own.
Use the same cue every time.
“After I pour my coffee, I write one line.” If-then plans had a meaningful effect on follow-through in a 2006 meta-analysis (Gollwitzer and Sheeran).
Give yourself room.
In our data, habits with a few-times-a-week goal were still going in week 4 more often than daily ones (11.4% vs 7.1%).
Make it enjoyable.
Pair it with something you like: a podcast on the walk, a nice pen for the journal.
Don't rely on reminders alone.
Habits with a reminder set did no better by week 4 (7.9% vs 7.7%). A notification is easy to swipe away.
Do it with one or two people.
6.6 check-in days in the first month with 1-2 friends, against 1.5 alone. In a 1999 weight-loss study, 95% of people recruited with friends and given social support finished the program, against 76% recruited alone (Wing and Jeffery).
Alone
3.1%
2 people
12.7%
3 people
26.4%
4-5 people
10.9%
6 or more
3.1%
Source: The Daily Pact Habits Report 2026.
See the numbers+
| People | Habits (n) | Alive in week 4 | Check-in days, month 1 |
|---|---|---|---|
| Alone | 1,615 | 3.1% | 1.5 |
| 2 people | 766 | 12.7% | 5.4 |
| 3 people | 444 | 26.4% | 8.9 |
| 4-5 people | 236 | 10.9% | 4.7 |
| 6 or more | 1,343 | 3.1% | 4.8 |
A few times a week
11.4%
Every day
7.1%
Reminder set
7.9%
No reminder
7.7%
n = 3,784 daily and 620 weekly-target habits; 520 with a reminder, 3,884 without. Source: The Daily Pact Habits Report 2026.
Get through the first two weeks together
Daily Pact is a habit tracker for you and one or two friends: one tap a day, and your pact sees when you've done it. That's the part of habit formation willpower can't do for you.
07A plan
A 66-day plan that survives week two
Plan the habit in three stages. The goal of the first one is not automaticity, it's survival.
Days 1-14
Survive
Pick one small habit and one fixed cue. Start with a friend, check in daily, and react to each other's check-ins. This is where 86% of stopped habits end.
Days 15-30
Protect
Expect a bad day. Never miss twice. If life gets in the way, switch to a few times a week instead of stopping.
Days 31-66
Keep going
Don't stop when a 30-day challenge ends. Same habit, same cue. Day 66 of 2027 is March 7, a better finish line than January 31.
Check-in days in the first month
1.5 days
alone
6.6 days
with 1-2 friends
Start on the 1st of a month
6.5%
other days
17%
the 1st
Still going in week 4. More in the fresh start effect.
Want a ready-made start? Try the January challenge, a printable habit tracker, or read Tiny Habits vs Atomic Habits for the method behind small habits.
FAQ
Questions people ask
How long does it take to form a habit?+
About two months on average, with a huge spread. In the best-known study (Lally et al., 2010) the median was 66 days for a daily behavior to become automatic, and individuals ranged from 18 to 254 days. A 2024 review of 20 studies found medians of 59 to 66 days. Simple habits get there faster than effortful ones.
Is it 21 days or 66 days to form a habit?+
66 is closer to the research. The 21-day figure comes from a 1960 book by plastic surgeon Maxwell Maltz, who wrote that patients needed a minimum of about 21 days to adjust to a change. It was an observation, not a habit study, and the 'minimum of about' got lost over the years.
How many days does it take to form a habit in real life?+
Most real habits never get the chance. In Daily Pact data on 4,404 habits, 86% of the habits that stopped did so in the first two weeks, and only 0.2% of daily habits reached a 66-day streak. The practical question is not 'how many days' but 'how do I get through days 7 to 14'.
What is the 21/90 rule?+
A popular idea that it takes 21 days to build a habit and 90 days to make it a lifestyle. Neither number comes from research. The useful part is the shape: a short first stage, then a much longer stretch before the behavior runs on its own.
Does missing a day reset habit formation?+
No. In the Lally study a single missed day did not materially affect progress. Our data shows what matters is the second miss: after one missed day, 28% of shared habits were checked in the next day; after two missed days in a row, only 15% were. Never miss twice: the data
How can I form a habit faster?+
Keep it small, tie it to the same cue every day, and do it with someone. Habits shared with one or two friends got 6.6 check-in days in the first month against 1.5 alone, and groups of three had the most habits still going in week 4 (26%).
How long does it take to form an exercise habit?+
Longer than for simple habits. Lally's participants took longer to automate effortful behaviors like sit-ups than drinking water. In our data, 91% of fitness habits shared with friends got a check-in in their first week, against 36% of solo fitness habits.
Sources
- The Daily Pact Habits Report 2026: 4,404 habits by 1,605 people in 86 countries, March to November 2026. App data, not a controlled trial.
- Lally et al. (2010), How are habits formed: Modelling habit formation in the real world: 96 volunteers, median 66 days to automaticity, range 18 to 254
- Singh et al. (2024), Time to form a habit: a systematic review and meta-analysis: 20 studies, medians of about 59 to 66 days, range about 4 to 335
- Gollwitzer and Sheeran (2006), Implementation intentions and goal achievement: meta-analysis of if-then plans
- Wing and Jeffery (1999), Benefits of recruiting participants with friends: 166 people; 95% of those recruited with friends plus social support completed the program vs 76% recruited alone with standard treatment
Get to day 66 with a friend
Start your 14-day free trial. No credit card required.
5.0 from 13 App Store ratingsAvailable on iPhone & Android · 14 days free for new accounts