How Long Does It Take to Form an Eating Habit?

About two months on average, with a range so wide the average barely helps#

How long does it take for a new eating habit to stick? The most-cited study puts the median at 66 days, but individual times ranged from 18 to 254 days¹. A 2024 systematic review of 20 studies found a similar pattern: in the four studies that reported formation times, separate study medians of 59–66 days and individual times from 4 to 335 days². So the fair answer is "about two months for a typical person, and possibly much longer for you." That isn't a failure; the spread is part of the finding.

Two findings matter more than the number. First, missing a single day didn't materially derail the process. Second, habits don't form in a straight line. The early repetitions build automaticity fastest, and then progress levels off. Both are worth keeping in mind for the first few weeks of any new eating routine.

That makes the 66 days a study median across different behaviors and people, not a deadline. If day 67 arrives and the habit still takes effort, you're not behind. You may just be on the slower side of a very wide range.

Where 66 days comes from#

Phillippa Lally and colleagues at University College London asked 96 volunteers to choose one eating, drinking, or activity behavior and do it daily in the same context for 12 weeks. The chosen behaviors covered healthy eating, drinking water, and exercise. Every day, participants rated how automatic the behavior felt using a standard questionnaire, the Self-Report Habit Index.

The researchers then fitted a curve to each person's ratings. Their main finding was that "automaticity increases following an asymptotic curve which can be modelled at the individual level"¹: steep at first, then leveling off at a plateau. The modeled time for each person to reach 95% of their own plateau is where the median of 66 days comes from. It was based on the 39 of 82 participants whose data fitted the curve well, and because data were collected over 84 days, many of those times were projected beyond the study².

Three details from the study rarely make it into the headline:

The range was enormous: 18 to 254 days, including modeled times well past the end of data collection.
Behaviors differed, but inconclusively. Drinking behaviors were performed more consistently, and the exercise group took about 1.5 times as long, but differences in the curves between behavior groups were not statistically significant in this small sample.
Early repetitions counted most. Each early repetition raised automaticity more than later ones, until more repetition stopped adding anything.

Missing a day doesn't reset the clock#

For anyone who treats a broken streak as a reason to quit, this is the most useful finding. In Lally's data, "missing one opportunity to perform the behaviour did not materially affect the habit formation process"¹. A 2012 review by the same group puts it plainly for clinicians: "missing the occasional opportunity to perform the behaviour did not seriously impair the habit formation process"³.

The streak mindset gets this wrong. A streak counter treats 60 unbroken days as fundamentally different from 29 days, one miss, and another 30. Habit formation doesn't work that way: what builds up is the link between the cue and the action, and one miss barely weakens it. The thinking that turns one miss into a lost week is the subject of our piece on all-or-nothing dieting.

This finding covers occasional misses. The study wasn't designed to show that frequent misses are harmless. Repetition is still the mechanism.

A missed day barely registers in the habit data. Treating it as a reset may be the bigger risk.

The "21 days" figure is folklore#

The belief that a habit takes 21 days is everywhere, and it appears not to come from habit research. According to Gardner and colleagues, the myth "appears to have originated from anecdotal evidence of patients who had received plastic surgery treatment and typically adjusted psychologically to their new appearance within 21 days"³. Getting used to a new face is a different process from making a new behavior automatic.

The cost of the myth is practical. If you expect a habit in three weeks and it still feels effortful on day 22, it's easy to decide you've failed or aren't "a habit person." Reviews of the real data advise expecting about 10 weeks. For some people and some behaviors, it takes much longer.

Claim	What the evidence shows
"It takes 21 days to form a habit."	Appears to trace to plastic-surgery anecdotes. Study medians: about 59–66 days
"Everyone takes about the same time."	Individual range of 18–254 days in one study, 4–335 days across studies
"Miss a day and you start over."	One missed opportunity did not materially affect formation
"Once it's a habit, it stays forever."	Habits can persist after motivation fades, but they depend on the cue remaining

What automaticity means for eating and logging#

"Automatic" in this research has a specific meaning. It doesn't mean effortless or permanent. It means the behavior is triggered by a cue in your environment rather than by a conscious decision. Gardner and colleagues define habits as "actions that are triggered automatically in response to contextual cues that have been associated with their performance," and a cue can be an event ("when I get to work") or a time of day ("after breakfast")³.

That has three practical consequences for eating routines:

1. Tie each new behavior to a cue that already happens every day. "Eat more vegetables" has no cue. "When I serve dinner, half the plate goes to vegetables first" does. The same applies to logging meals: "log right after I sit down to eat" gives the behavior a trigger. Our guide to building a calorie-tracking habit goes further into choosing that trigger. Writing the cue and action as a single sentence is the approach covered in if-then plans.

2. Choose the behavior yourself, and consider timing. The 2024 review found that self-selected habits and morning practices generally became stronger. In one study it describes, stretching (not an eating behavior) took an average of 106 days to become a habit when done in the morning and 154 days in the evening². Whether the same holds for food habits hasn't been tested.

3. Expect the habit to depend on its setting. Because habits are tied to cues, a change of setting such as travel, a new job, or a holiday can disrupt them. That is not a sign the habit was fake. It needs to be re-anchored to new cues.

Does the habit approach help weight loss?#

Habit formation is a way of making behaviors last, and there is some direct evidence for weight. In a small exploratory trial, 104 adults with overweight or obesity were allocated (69 to the intervention, 35 to a waiting list); the intervention group received a leaflet with habit-formation advice, a short list of eating and activity tips, and a checklist for tracking whether they did them. After eight weeks they had lost 2.0 kg on average, against 0.4 kg on a waiting list. At 32 weeks, intervention participants still in the study had lost 3.8 kg; carrying forward the last recorded weight for dropouts gave 2.6 kg. The 32-week figures cover the intervention group only⁴.

Those are modest numbers from a simple, low-cost intervention. The larger point is that the behaviors that lead to weight loss can be made automatic. Once they are, they may depend less on motivation, which often fades. Gardner and colleagues note that habits "are likely to persist even after conscious motivation or interest dissipates"³. For how this fits into a broader plan, see our guide to calorie deficits and our guide to sustainable weight-loss habits.

FAQ#

Is it really 66 days to form a habit?#

Sixty-six days was the median in one well-known study, and people ranged from 18 to 254 days¹. A later review found medians of 59–66 days, with a very wide spread between people². Use it as a rough expectation, not a deadline.

Where did the 21-day habit idea come from?#

From anecdotes about plastic-surgery patients who adjusted to their new appearance in about 21 days, not from studies of behavior change³.

Which eating habits are easiest to form?#

Simple ones tied to an existing daily cue. In Lally's study, drinking behaviors were performed more consistently, though differences between behavior types were inconclusive, and a review found that self-chosen habits and morning routines generally became stronger².

Sources#

Source: BurnWeek — "How long does it take to form a healthy eating habit?", https://burnweek.fit/blog/how-long-to-form-an-eating-habit/. Licensed CC BY 4.0: free to quote or reuse with a link to this page.

This article was researched and drafted with AI assistance and reviewed for accuracy by the BurnWeek team. It is general information, not medical advice. How we research and correct our articles →