Chapter 4: The Truth About Twenty-One Days

Day Twenty-Two

Nadège is thirty-eight and works as a receptionist at a dental clinic. At the start of the year, she decided to build a new habit: writing in a journal for ten minutes every evening before bed. She had read that writing about her day might help her feel calmer and sleep better, and she liked the idea. She had also read, in several places, that it takes twenty-one days to form a habit. So she bought a beautiful notebook, drew a chart with twenty-one boxes and began.

For the first week, journaling was easy. Nadège was enthusiastic, and she enjoyed writing. In the second week, it became a little harder. Some evenings she was tired, and she wrote only a few lines. In the third week, it felt like a chore. She kept going because of the chart, ticking off each box. On the twenty-first evening, she ticked the last box and felt proud. She had done it. The habit was formed.

On the twenty-second evening, Nadège was tired after a long day, and she went to bed without writing. On the twenty-third evening, she remembered the journal only after she had turned off the light. By the end of the month, she had not written for a week. The habit she thought she had formed had simply disappeared. She felt confused and disappointed. She had done the twenty-one days. Why had it not worked?

When Nadège mentioned this to her sister, who works as a researcher in a university, her sister laughed kindly. "Twenty-one days is a myth," she said. "It seems to come from an old self-help book, not from research on habits. It usually takes much longer than that, and it varies enormously." Nadège was surprised. She had seen the number so often that she had assumed it must be true.

The idea that a habit takes twenty-one days to form is one of the most widely repeated claims about habits. It appears in books, articles and advertisements. But when scientists have actually measured how long it takes for a new behaviour to become automatic, they have found a very different and much more interesting picture. This chapter looks at what they found, and what it means for anyone trying to build a new habit.

Where the Myth Came From

The number twenty-one seems to have entered popular culture through a self-help book published in the mid-twentieth century. Its author was a surgeon who had noticed that his patients often took around three weeks, at least, to get used to changes in their bodies or their lives, such as a change in their appearance after an operation. He wrote that it seemed to take a minimum of about twenty-one days for an old mental image to fade and a new one to take hold.

This observation was about adjusting to a change, not about forming a new habit, and the surgeon described it as a minimum, not a fixed rule. But over the following decades, as the idea was repeated and simplified, "a minimum of about twenty-one days to adjust" became "twenty-one days to form a habit". The number was easy to remember, and it promised a clear finish line. It spread widely, but it was never based on research into habit formation.

This is a common pattern in popular ideas about the mind and brain. An observation made in one context is simplified, repeated and stretched to cover something else. Each time it is repeated, it seems more authoritative, because so many people have said it. But repetition is not evidence. When we hear a precise number about how the mind works, it is often worth asking where it came from.

The twenty-one-day myth is not only inaccurate. It can be harmful, as Nadège found. People who believe it may expect a habit to be fully formed after three weeks, stop paying attention and then be surprised and discouraged when the habit fades. They may conclude that they are weak or that habits do not work for them, when in fact they simply stopped too early.

There is also a quieter harm. A fixed number encourages us to think of habit formation as a test with a pass mark, rather than as a gradual process. People start counting days instead of paying attention to how the behaviour is fitting into their lives. They focus on the chart instead of the cue, the reward and the situation. The twenty-one-day idea is attractive precisely because it is simple. But habits are not simple in that way, and treating them as if they were often leads to disappointment.

So what does the research actually show? The answer comes from studies in which researchers followed people as they tried to build new habits, measuring how automatic the habits felt over time.

What Researchers Found

One often-cited study asked a group of volunteers to choose a simple new behaviour to do every day in the same situation. Some chose to eat a piece of fruit with lunch, some to drink a glass of water after breakfast, some to do a short exercise before dinner. Each day, the volunteers reported how automatic the behaviour felt, using questions such as whether they did it without thinking and whether it would feel strange not to do it. The researchers followed them for twelve weeks and used the results to estimate how long it took for each person's behaviour to reach its highest level of automaticity.

The typical time was about 66 days, a little more than two months and much longer than twenty-one days. But the most striking result was how much people differed. For some people, a behaviour became automatic in about 18 days. For others, the estimated time stretched to about 254 days, more than eight months; times longer than the twelve weeks of the study were estimates, worked out from how each person's progress was rising. Simple behaviours, like drinking a glass of water, tended to become automatic faster than more demanding ones, like exercising. And some people's behaviours never reached high automaticity during the study at all.

The study also found something encouraging. Missing a single day did not seriously damage the process. People who skipped the behaviour once in a while still went on to build their habits. What seemed to matter was repeating the behaviour consistently, in the same situation, over a long period, not achieving a perfect unbroken record.

Another important finding was the shape of the progress. Automaticity did not grow at an even rate. It grew fastest in the early days and weeks, and then more slowly, gradually levelling off, like a curve that rises steeply at first and then flattens. In other words, the first repetitions do a great deal of the work, and later repetitions add smaller and smaller gains until the habit is as automatic as it will become.

It is important to say that this was one study, with a limited number of volunteers, and that other research has found different times for different behaviours and different people. Scientists do not have a single number for how long habits take to form, and they probably never will, because it depends on the person, the behaviour, the situation and how consistently it is repeated. But the general picture, that it usually takes longer than three weeks, varies widely and does not require perfection, is supported by research and is much more useful than the myth.

Nadège Starts Again

After talking with her sister, Nadège decided to start her journaling habit again, with a different approach. She threw away the chart with twenty-one boxes. Instead, she decided to give the habit at least three months, and to focus on consistency rather than perfection. If she missed a day, she would simply write the next day, without feeling that she had failed.

She also thought about the situation in which she wrote. In her first attempt, she had written at different times and in different places: sometimes in bed, sometimes at the kitchen table, sometimes on the sofa. She realized that this made it harder for a habit to form, because there was no consistent cue. So she chose one situation: after brushing her teeth, sitting on the edge of her bed, with the journal and a pen kept on her bedside table.

She made the habit smaller, too. Instead of ten minutes, she committed to writing just three lines each evening. If she felt like writing more, she could. But three lines were enough to count. On tired evenings, three lines took less than a minute, so she rarely had a reason to skip.

Over the following months, Nadège noticed the curve her sister had described. In the first few weeks, she still had to remember to write, and sometimes she forgot. By the second month, she found herself reaching for the journal almost automatically after brushing her teeth. By the third month, it felt strange not to write. She had missed perhaps eight or nine evenings over the three months, but the habit had formed anyway.

"The twenty-one days made me think I had a finish line," Nadège says. "When I crossed it, I stopped paying attention, and the habit fell apart. The second time, I knew it would take longer and that missing a day was fine. That took all the pressure off." Nadège's story shows how much a realistic expectation can help. Knowing that habits take time, and that perfection is not needed, made it easier for her to keep going until the habit was truly built.

Bilal and the Long Road to Running

Bilal is twenty-six and works as a delivery driver. He decided to start running three times a week, after work, to improve his fitness. He had heard that it would become a habit in a few weeks. But after two months, he still found it hard to make himself go. Every run felt like a decision he had to make, and on many evenings he chose not to.

Bilal's experience fits what the research suggests about more demanding behaviours. Running is not like drinking a glass of water. It takes effort, time and preparation, and it can be uncomfortable, especially at the start. Such behaviours tend to take longer to become automatic, and for some people they may never become fully automatic in the way that brushing teeth does. That does not mean they cannot become a regular part of life. It means they may always need a little more planning and support.

Bilal decided to give himself more time and to make the habit easier. He chose a consistent cue: arriving home from work on Monday, Wednesday and Friday, changing straight into his running clothes before sitting down. He kept his running clothes laid out on a chair by the door. He started with shorter runs, sometimes only fifteen minutes. And he stopped expecting running to feel automatic soon.

It took Bilal about five months before running began to feel like a natural part of his week rather than a constant decision. Even then, he says, there are days when he has to push himself to go. But the push is much smaller than it was. "I thought something was wrong with me because it did not become easy after a few weeks," he says. "Now I know that running is just a harder habit to build. It took longer, and that is normal." Anyone starting a new exercise routine should check with a doctor or another qualified professional about what is suitable for them.

Bilal's story is a reminder that not all habits are equal. Some behaviours become automatic quickly. Others take many months. Some may never become fully automatic, but can still become reliable through good planning, consistent cues and patience. Comparing yourself with someone who built a different habit, or who is a different person, can be discouraging and misleading.

Wanda's Morning Stretch

Wanda is seventy and lives alone in a small flat. Her doctor suggested some gentle stretches each morning to help with stiffness in her back and shoulders, and showed her five simple movements that would take about five minutes. Wanda was doubtful that she would remember to do them every day. She had never been good at keeping routines, she said.

Wanda decided to attach the stretches to something she already did every morning without fail: making her first cup of tea. While the kettle boiled, she would do her five stretches in the kitchen. She stuck a small card with drawings of the stretches on the kitchen cupboard, next to the tea caddy, as a reminder.

To her surprise, the habit formed quickly. Within about three weeks, Wanda found herself starting the stretches as soon as she switched on the kettle, without needing to look at the card. The behaviour was simple, the cue was strong and consistent, and the stretches fitted neatly into the few minutes she was already spending in the kitchen. After a couple of months, she noticed that her back felt less stiff in the mornings.

Wanda's experience shows the other end of the range. For her, a simple behaviour, attached to a strong existing routine, became automatic within a few weeks. Her experience was not better or worse than Bilal's. It was simply different, because her habit was simpler and its cue was very strong. The research suggests exactly this kind of variation.

"I always thought I was bad at habits," Wanda says. "But it turns out I was just choosing the wrong ones, or trying to do them at the wrong time. When I tied the stretches to my tea, it was easy." Her story suggests that simple behaviours, attached to strong existing cues, are often the easiest way to begin building new habits, and can build confidence for harder ones.

Kasimir and the Train

Kasimir is forty-four and works as a quality controller in a factory that makes car parts. He has always wanted to learn another language, and he decided to study for ten minutes every morning on the train to work, using a small book of vocabulary and short exercises. The train journey was the same every weekday, at the same time, in the same kind of seat, which gave him a very consistent context.

For the first few weeks, Kasimir had to remind himself to take out the book. But because the context was so consistent, the habit grew steadily. By the second month, he found that he reached for the book as soon as he sat down on the train. His study became part of the journey, as automatic as looking out of the window.

Then came the summer holidays. For three weeks, Kasimir did not take the train, and he did not study at all. When he returned to work, he worried that the habit would be lost. But on his first morning back, as he sat down on the train, his hand reached for the book almost by itself. The habit had survived the break, because the cue, the train seat, was still linked with the behaviour.

Kasimir did notice, however, that he had not studied at all on holiday, even though he had plenty of free time. Without the train, there was no cue, and the habit did not appear. This illustrates how closely habits are tied to their context. The habit was not "studying every day". It was "studying on the morning train". In a different context, it did not run.

Kasimir decided that this was fine. His habit served his everyday life well, and he did not need it to work on holiday. But for his next holiday, he planned a small new cue: studying for ten minutes with his morning coffee, wherever he was. "I learned that a habit lives in a place," he says. "If you change the place, you need a new cue." His story shows both the strength of consistent contexts and the reason why habits sometimes disappear when our routines change.

Practical Steps: Giving Habits the Time They Need

The first step is to let go of the twenty-one-day finish line. Expect a new habit to take weeks or months to become automatic, depending on how simple it is and how consistent its context is. Do not stop paying attention after three weeks. Keep going until the habit feels natural, and even then, keep an eye on it for a while.

The second step is to choose a consistent context. Do the new behaviour in the same situation each time: after the same existing routine, at the same time, in the same place. Consistency of context is one of the most important factors in how quickly a behaviour becomes automatic.

The third step is to start simple. If you are building a new habit, especially your first, choose something simple and easy to repeat. Simple habits tend to form faster, and success with them builds confidence. Harder habits can come later, or can be broken into simpler parts.

The fourth step is to aim for consistency, not perfection. Missing a day now and then does not seem to seriously damage habit formation. What matters is returning to the behaviour as soon as you can. Do not let a missed day become a missed week because you feel you have failed.

The fifth step is to notice the curve. Pay attention to how automatic the behaviour feels over time. You may notice that it becomes easier quickly at first, and then more slowly. This is normal. The early weeks do a lot of the work, and the later weeks consolidate it.

The sixth step is to expect differences. Your habit may take longer or less time than someone else's, and one of your habits may take longer than another. This variation is normal and does not mean that you are doing something wrong. Compare yourself only with your own progress.

Automatic Is Not the Same as Permanent

Even when a habit has become automatic, it is not necessarily permanent. Habits are tied to their contexts, as Kasimir found. When contexts change, through holidays, illness, moving house or a new job, habits can weaken or disappear. This is not a failure. It is how habits work. When your life changes, it may be necessary to rebuild some habits in the new context, often with a new cue.

It is also worth saying that not every behaviour needs to become fully automatic to be valuable. Some important behaviours, such as exercise for many people, may always require some effort and decision. They can still become regular and reliable parts of life through good planning, consistent cues and support. If a habit never feels as automatic as brushing your teeth, that does not mean it has failed.

Finally, the research described in this chapter concerns ordinary, simple habits. Changing deeply established patterns, especially those connected with health, mood or addiction, can be much more complex and may take much longer. If you are trying to change such a pattern and finding it very hard, please consider seeking support from a doctor or another qualified professional. Their help can make a real difference.

A Short Exercise: The Automaticity Diary

This exercise helps you follow the growth of a new habit over time, in a way that is closer to the research than a simple tick chart. It takes less than a minute each day, for as long as you wish to continue.

Choose one simple new habit you would like to build, and decide on a consistent context: after which existing routine, at what time and in what place you will do it. Write this down at the top of a page. For example: "After brushing my teeth at night, sitting on the bed, write three lines in my journal."

Each day, after doing the behaviour, or at the end of the day if you did not, write the date and a number from one to five answering this question: "How automatic did it feel today?" One means "I had to remember and push myself." Five means "I did it without thinking; it would have felt strange not to." If you missed the behaviour, write "missed" and move on.

Every two weeks, look back at your numbers. Notice whether they are rising, how quickly and whether missed days made any lasting difference. Do not expect a straight line. Expect a curve that rises faster at first and then levels off, with some ups and downs along the way.

Continue for at least two months, or until your numbers stay at four or five for several weeks. Nadège did this exercise during her second attempt at journaling. Her first two weeks were mostly ones and twos. By the sixth week, they were mostly threes and fours. By the eleventh week, they were almost all fives. She had missed nine days in total. "Seeing the numbers rise," she says, "made me trust that it was working, even on days when it felt hard."

Chapter Summary

The popular claim that it takes twenty-one days to form a habit appears to come from a mid-twentieth-century self-help book in which a surgeon observed that people took at least about three weeks to adjust to changes; it was never based on research into habits. One often-cited study found that it took volunteers about 66 days, a little more than two months, for a new daily behaviour to become as automatic as it would get, with enormous differences between people, from about 18 days to about 254 days. Simple behaviours tended to become automatic faster than demanding ones, progress was fastest at the start and missing an occasional day did not seriously harm the process.

Nadège and her journal, Bilal and his running, Wanda and her morning stretches and Kasimir on his train all found that realistic expectations, consistent contexts, simple beginnings and consistency rather than perfection made habits easier to build. Automatic is not the same as permanent: habits are tied to their contexts and may need rebuilding when life changes. Some behaviours may never feel fully automatic, and that is not failure.

In the next chapter, we will look at another popular idea about habits: that changing them is mainly a matter of willpower. The science of willpower turns out to be more complicated, and more hopeful, than many people think.