Every habit book on the shelf tells some version of the same origin story. You repeat a cue-routine-reward loop enough times, the groove gets deeper, and one day — somewhere between 18 and 254 days, if you believe the oft-cited University College London number — the behavior stops requiring willpower. Slow, gradual, incremental. That’s the entire premise underneath the daily-repetition programs in Atomic Habits, The Power of Habit, and most of what gets written about behavior change.

A study published in Nature Communications on June 3, 2026 says that story is wrong. Not oversimplified. Wrong about the mechanism.

Quick Verdict

Question	Answer
Who ran the study?	Sharlen Moore, Kishore V. Kuchibhotla, and colleagues at Johns Hopkins University
Where was it published?	Nature Communications, June 3, 2026 (DOI: 10.1038/s41467-026-71048-0)
What did they test?	Whether the shift from deliberate to habitual behavior in mice happens gradually or abruptly
What did they find?	Abruptly — a switch-like transition, not a slow slope
What’s driving it?	A shift in the dorsolateral striatum from outcome-tracking to stimulus-response processing
Does this kill the habit loop?	It contradicts the mechanism, not necessarily the advice — see below

Bottom line: The century-old assumption that habits build up slowly through repetition — the exact premise habit books use to sell daily-practice systems — doesn’t match how the transition actually happens in the brain.

What the Study Actually Tested

Researchers gave mice a go/no-go task, then let their home cages carry citric-acid water that reduced (without eliminating) their motivation to seek out the good stuff. The point was to watch, trial by trial, exactly when a mouse’s behavior stopped tracking the outcome it wanted and started running on autopilot instead — the same goal-directed-to-habitual shift habit books describe in humans, just measurable in an animal that can’t self-report it.

The question goes back further than James Clear or Charles Duhigg. Psychologist Clark Hull was already arguing about whether habits form slowly or suddenly back in 1943. Eighty-three years of psychology mostly assumed “slowly.” This is one of the first studies with the tools to actually watch the transition happen and check.

Do Habits Form Gradually or Suddenly?

Habits form suddenly, not gradually. Using statistical modeling on trial-by-trial behavior, the Johns Hopkins team pinpointed the shift from goal-directed to habitual action to roughly a three-trial window — a switch flipping, not a dial turning. Three findings carry the argument:

1. The behavioral switch was abrupt, not sloped. Fluctuations in the mice’s responses continued well after their performance on the task had already stabilized, then stopped all at once. A gradual-reinforcement model predicts a smooth decline in those fluctuations. That’s not what happened.
2. The brain data matched the behavior data. Fiber photometry recordings of the dorsolateral striatum showed outcome-related activity drop and stimulus-response activity sharpen at the same moment the behavioral switch occurred — not a slow crossover, a coincident flip.
3. A specific brain region appears to gate the transition. The dorsolateral striatum’s shift looks less like accumulated wear and more like a controller getting recruited. Senior author Kishore Kuchibhotla told Johns Hopkins’ Hub that identifying a possible controller “means maybe we can reverse maladaptive habits back to goal-directed behavior” — which is a very different claim than “habits are grooves that only get deeper.”

Why This Is a Problem for the Habit Loop

The cue-routine-reward loop that anchors the habit-formation shelf isn’t just a description of what a habit looks like once it exists. It’s a claim about how it gets built — incremental reinforcement, cue paired with reward, over and over, until the behavior automates. That’s the model behind habit stacking, behind 66-day timelines, behind the entire “show up daily and the automaticity will come” pitch.

This study says the actual transition in the brain doesn’t look like that. It looks like a threshold getting crossed, and then a different neural system taking over almost immediately. Kuchibhotla’s team is careful about what this does and doesn’t overturn — mice aren’t humans, and the study measured a simple auditory task, not “flossing” or “going to the gym.” But it’s a direct test of a mechanism the self-help industry treats as textbook, and the mechanism didn’t hold up the way the textbook says.

We’ve flagged this exact pattern before: the genre borrows a real scientific concept, states it more confidently than the research supports, and builds a program on top of the overstatement. We saw it with digital reminders and cognitive offloading — a “no downside” claim that a direct experiment simply didn’t support. We saw it with self-help products and personality change — a category-wide null result on the outcomes the industry sells hardest. Now it’s the mechanism underneath the single most successful subgenre in self-help: habit formation itself.

What This Doesn’t Mean

This isn’t a case for throwing out your habit tracker. A few things the study doesn’t say:

• It doesn’t say the practical advice fails. Environment design, habit stacking, and small starting behaviors can still work as inputs — the study challenges the claimed mechanism of gradual reinforcement, not whether repetition and structure help you get somewhere.
• It doesn’t say mice generalize perfectly to people. This was a simple go/no-go task in a controlled lab setting, not a longitudinal study of humans trying to floss more.
• It doesn’t say every habit forms in three trials. The “abrupt” finding describes the shape of the transition — a switch, not a slope — not a universal three-repetition rule for locking a new behavior into place.

What it does say: the story habit books tell about why their systems work — slow accumulation, deepening grooves, neural pathways strengthening with each rep — isn’t how the transition actually happens in the brain, at least in the best-controlled test of it so far.

The 66-Day Number Was Already Shaky

Even before this study, the specific timelines habit books cite were softer than the marketing suggests. The famous “66 days to form a habit” figure comes from a 2010 UCL study on 96 people, and the real range in that data ran from 18 to 254 days — a spread so wide it barely functions as a number. Habit books tend to quote the 66 in headlines and bury the 18-to-254 in a footnote, if it shows up at all.

The Johns Hopkins findings don’t fix that range problem or replace it with a cleaner one. What they do is explain, mechanistically, why a clean number was never going to exist. If the transition is a threshold that gets crossed rather than a slope that gets climbed, then “how many days” was always the wrong unit. The better question is what determines when the threshold gets crossed — and that’s a controller in the brain, not a calendar.

Why This Matters for a Site Built on Habit Books

Four posts on this site — the Atomic Habits review, the Atomic Habits vs. Power of Habit comparison, the Atomic Habits vs. Tiny Habits comparison, and the Atomic Habits workbook review — evaluate these books on implementation: what stuck, what didn’t, whether the framework survives contact with an actual life. None of them stopped to ask whether the underlying science the books lean on for their credibility — “this is backed by behavioral research” — was itself settled. It wasn’t, and this study is a decent illustration of why: the field’s best answer to “how do habits form” changed substantially in 2026, and the popular books were written years before that.

That’s not a reason to retract the practical verdicts in those reviews. Environment design and habit stacking can be useful tools regardless of whether the underlying mechanism is a slope or a switch — a screwdriver still turns a screw even if your theory of torque is off. But it’s a reason to hold the scientific authority claims in those books more loosely than the confident tone suggests.

What Actually Holds Up, Given This

If the mechanism is a threshold rather than a slope, a few practical implications follow:

• Don’t treat a slow start as proof the habit “isn’t sticking.” If the switch is abrupt, a habit that feels effortful on day 40 could flip on day 41. Gradual visible progress was never guaranteed by the old model either — this just explains why it doesn’t show up.
• Consistency still matters, but the “compounding” framing might be the fiction, not the consistency. We flagged the compounding-math problem in Atomic Habits directly — 1% daily gains don’t compound the way the famous chart implies. This study adds a mechanistic reason: there’s no smooth curve to compound along.
• If a habit feels genuinely stuck, more repetition of the same routine may not be the lever. Kuchibhotla’s framing of a “controller” suggests the more promising target for stubborn or unwanted habits is whatever gates the switch — not just doing the behavior more times and hoping the groove deepens.

The Bottom Line

A well-designed Johns Hopkins study, using brain recordings alongside trial-by-trial behavioral tracking, found that the shift from deliberate to habitual behavior happens abruptly — a switch, not a slope — governed by a specific, identifiable shift in the striatum. That contradicts the slow-reinforcement mechanism that Atomic Habits, The Power of Habit, and most habit-formation self-help use to justify daily-repetition programs. The practical advice in those books isn’t necessarily wrong. The story about why it works was never tested this directly before — and now that it has been, the “deepening groove” isn’t what the data shows.

---

Revealing abrupt transitions from goal-directed to habitual behavior, Moore, Kuchibhotla, et al., Nature Communications, June 3, 2026 (DOI: 10.1038/s41467-026-71048-0). Additional coverage from Johns Hopkins Hub and Futurity. Related reading: Atomic Habits Review, Atomic Habits vs The Power of Habit, Your Reminder App Is Quietly Wrecking Your Memory, and Self-Help Products Don’t Change Your Personality: Study.