How long does it take to break a habit? Not 21 days

There is no fixed number of days. The study everyone quotes, by Phillippa Lally and colleagues at University College London, measured how long it takes to form a new habit, not to break an old one, and found a median of 66 days with a range from 18 to 254. The popular 21 days is not a finding at all, and the research on breaking habits is thinner still.

So I will not promise you a date. What I ask for instead is thirty days, as a test you judge for yourself. Here is the reasoning.

Where the 21 days came from#

In 1960 a plastic surgeon named Maxwell Maltz published Psycho-Cybernetics, a book about self-image. He had noticed that patients seemed to take about three weeks to get used to a changed face, and he saw similar adjustments elsewhere, like starting to feel at home in a new house. He wrote that it takes "a minimum of about 21 days" for an old mental image to give way to a new one.

Two things happened as the line traveled. The "minimum" fell off. And getting used to something turned into forming a habit. Habit researchers at University College London have suggested that this second slide, confusing habituation with habit formation, is how a surgeon's anecdote became a rule. It was an anecdote, not a measurement.

What the research measured#

Lally's study, published in 2010, asked 96 volunteers to pick one eating, drinking or activity behavior and do it every day in the same context, for example after breakfast, for twelve weeks. Each day they rated how automatic it felt. For the people whose data could be modeled well, the median time to reach 95 percent of their eventual plateau was 66 days. The range ran from 18 to 254 days, and the longest figures were projections beyond the 84 days the study actually observed. Simpler actions, like drinking a glass of water, became automatic sooner than harder ones.

A systematic review in 2024 gathered 20 studies with 2,601 people. Only four reported how long habits took to form: medians of 59 to 66 days, means of 106 to 154 days, and individuals ranging from 4 to 335 days. Most of the studies carried a high risk of bias, so treat every one of these numbers as rough.

Notice what all of it is about: building small, healthy habits like flossing, stretching or drinking water. None of these studies asked people to stop something compulsive.

James Clear makes a related point in Atomic Habits: habits form through repetitions, not through the passing of days. A calendar does not build a habit, and it does not unbuild one either.

Breaking a habit is a different question#

You cannot form a habit of not doing something. Lally and her colleagues made that point themselves, and added that breaking an existing habit takes different and more effortful strategies than making a new one. The ones I found most useful come down to changing the setting the habit runs on, which is what how to break a bad habit is about.

On how habits actually break, the evidence is much thinner. In 2023, nineteen habit researchers published twenty-one open questions for their field. What disrupting a habit does to the underlying link between cue and response, they wrote, "has received little empirical attention". Some expect the link to fade with long disuse. Others expect it to persist quietly, ready to be set off again by the old setting or by a stressful stretch of life.

I plan around the second view. If the old link can wake up again, then the changes that removed the cues are not training wheels you take off after a month. The phone keeps sleeping outside the bedroom. The filter stays on. That is why the rule you leave the path with keeps them in place after the count stops mattering.

What the first weeks are like#

I have not found a study that times this for pornography. The nearest thing is a psychiatrist's clinical experience. Anna Lembke, who teaches psychiatry and addiction medicine at Stanford and wrote Dopamine Nation, says she has asked patients for twenty years to stop the thing they overuse for a month. She describes them typically feeling worse for the first ten to fourteen days, and by week three or four reporting that they feel better than they have in a long time. That is an observation from her practice, not a measurement, but it is the most honest map I know of.

The lessons on the path describe the same shape from the inside. The first seventy-two hours are the loudest. The day after a slip can be louder still. And some weeks in, there may be a flat stretch that passes.

Why thirty days, then#

Not because something resets on day thirty. Because thirty days is long enough for the evidence to come in, and short enough to promise on day one.

The thirty days on this site are a test, not a cure with a timer. You stop completely. On day one you write one line about each of four things: how long you can stay with a hard task, what you are like with people, how you sleep, and where the hours go. On day thirty you read them again and judge. Knowing what the habit cost never changed anything for me. Taking it away for long enough to feel the difference did. That is the gap between knowing and realizing.

Lembke's month happens to match. She calls hers a kind of personal experiment, which is exactly how I would describe this one.

One more finding from Lally's study is worth carrying. Missing a single day did not materially set back the habit people were building. That is a finding about forming habits, not breaking them, so I hold it loosely. But it agrees with Clear's rule of thumb: one miss is an accident, and it is the second in a row that starts a new pattern. A slip is one event. Come back the same day.

Questions people ask#

Is it 21 days, 66 days or 90 days?#

None of them is a rule. Twenty-one began as a surgeon's anecdote, 66 is a median from one study of simple healthy habits, and I have not found a study behind the ninety days often quoted for pornography. Expect your own timing to vary, and judge by what changes rather than by the date.

Does missing a day reset a habit?#

Not in Lally's study, where a single missed day did not materially affect the habit being formed. For stopping, the spirit is the same: a slip is one event, and the days you kept going still happened.

Why do old habits come back after months?#

One view in habit research is that the link between a cue and a response does not vanish when you stop acting on it. It waits, and the old setting or a hard week can set it off. That is why the cues stay removed after the thirty days, not only during them.

Does it get easier with time?#

In Lembke's clinical experience, the first two weeks are usually the hardest and things improve after that. Removing the cues helps too, simply because fewer things are left to set the urge off.

If you would rather find your own answer than borrow a number, start the thirty days: stop completely, write the four lines on day one, and read them again on day thirty.