The Science of Making & Breaking Habits

High-Level Overview

In this solo episode, Andrew Huberman explores the biology and psychology of forming and breaking habits. He explains that habits are learned behaviors, not hard-wired reflexes, and that it's estimated "up to 70% of our waking behavior is made up of habitual behavior." Drawing on the neuroscience of plasticity and the research psychology literature, especially a review called Psychology of Habit by Wendy Wood and Dennis Rünger, he introduces practical concepts such as limbic friction, linchpin habits, habit strength and task bracketing, then builds them into two concrete programs: a three-phase daily schedule and a 21-day habit system. He closes with a counterintuitive, biology-based method for breaking bad habits.

The tone is methodical and practical, with a steady mix of mechanism (the basal ganglia, NMDA receptors, dopamine) and everyday examples like brushing teeth, zone 2 cardio, writing and reaching for the phone. The conversation moves from definitions to science to step-by-step tools, then ties everything together in a recap.

The Discussion Explores:

• Habits as learned neuroplasticity, and goal-based versus identity-based habits
• Why forming a habit can take anywhere from 18 to 254 days, and the idea of limbic friction
• Linchpin habits and how to measure habit strength
• Procedural memory visualization and task bracketing in the dorsal lateral striatum
• A three-phase daily plan that matches habits to the brain's natural chemistry
• Reward prediction error and how to use dopamine to support new habits
• A 21-day program of six habits a day followed by a 21-day test
• Breaking bad habits by adding a positive behavior right after the unwanted one

Deep Dive into Main Ideas

What Habits Are and Why They Take So Long to Form

Huberman distinguishes habits from hard-wired reflexes like blinking when something approaches your eye. Habits are learned, sometimes without our awareness, and learning is neuroplasticity: changes in the connections between neurons that make some behaviors more likely and others less likely.

He separates immediate goal-based habits, like getting a set amount of zone 2 cardio each week where you can check a box each time, from identity-based habits tied to a larger self-image, like becoming a fit person. He mentions the research recommending 150 to 180 minutes per week of zone 2 cardio, the kind of exercise where you can just barely hold a conversation.

As for how long habits take, he cites a 2010 study by Lally that tracked people trying to adopt walks after dinner, a behavior that helps clear glucose from the blood. Forming the habit, defined as doing it about 85 percent of the time with little mental effort, took anywhere from 18 to 254 days depending on the person. If one habit is hard for you, he notes, it doesn't mean all habits will be.

Limbic Friction and Linchpin Habits

To explain those differences, Huberman uses a term he coined: limbic friction. It describes the strain required to overcome one of two states, either being too anxious to calm down and do something, or being too tired and unmotivated to act. Both relate to the autonomic nervous system, which works like a seesaw between alertness and calm. Limbic friction is the activation energy needed for your forebrain to override your current state, and he prefers it to the vaguer notion of motivation because it points to specific circuits. High-energy habits are harder when you're tired, while calming habits like meditation are harder when you're keyed up.

Linchpin habits are habits you enjoy that make many other habits easier. For Huberman, those are resistance training and running, three or four times a week each, typically early in the day. Doing them makes it easier to be alert for work, sleep well, stay hydrated and choose better foods. He stresses that linchpin habits are always things we enjoy.

Measuring Habit Strength

From the psychology literature, habit strength depends on two things. The first is context dependence: whether you do the habit the same way regardless of setting. Huberman brushes his teeth at the same time relative to waking whether he's home or traveling, so that habit is context independent and strong. The second is how much limbic friction it takes to perform. The goal is automaticity: "Automaticity is fancy language for the neural circuits can perform it automatically."

Tool: Walk Through the Steps Once

The first tool comes from procedural memory, the memory for sequences of steps, as opposed to episodic memory for events. Simply thinking through, once, the specific sequence of steps needed to execute a new habit makes people much more likely to perform it, not just the first time but in the days and weeks after. His example is making an espresso: walk into the kitchen, turn on the machine, draw the shot. No elaborate meditation is required.

He explains the mechanism through Hebbian learning, from the Canadian psychologist Donald Hebb: neurons that are active together strengthen their connections. NMDA receptors play a key role, recruiting more receptors when a neuron gets strong input so it responds more easily in the future. Your nervous system knows the difference between thought and action, but rehearsing the steps sets the relevant neurons in motion so that "the dominoes fall more easily" when you actually do it. The same exercise can help you return to a habit you've dropped.

Task Bracketing and the Basal Ganglia

The second and, Huberman suggests, perhaps most powerful tool is task bracketing. The basal ganglia handle go (doing things) and no-go (not doing things) functions. Within them, a region called the dorsal lateral striatum becomes active at the beginning and end of a habit, not during it. It brackets the habit, creating a neural imprint that this thing happens at this point. Strong bracketing is what lets a habit happen even after bad sleep or on a hard day, like brushing your teeth no matter what.

He then challenges a common idea: scheduling a habit at an exact time helps in the short term but not in the long term, because the nervous system generates behavior based on state, not clock time. People who are rigid about exact times often don't stick with habits because of context dependence. It's better to anchor habits to phases of the day with predictable biology.

Program One: Three Phases of the Day

Using a typical schedule of bed around 10 PM and waking around 7 AM, Huberman divides the day into three phases.

Phase one, about 0 to 8 hours after waking, is naturally high in norepinephrine, epinephrine, dopamine and healthy morning cortisol. Morning sunlight (or bright artificial light) within 30 minutes of waking, exercise, cold exposure, caffeine, fasting, tyrosine-rich foods, and for those who use them supplements like alpha-GPC, phenylethylamine or L-tyrosine further support this alert, action-oriented state (he advises consulting a doctor about supplements). This is the time to place the habits with the highest limbic friction, the ones you most want but find hardest. It is not about getting unpleasant tasks out of the way; it's about using a phase when your whole system is oriented toward action.

Phase two, about 9 to 14 or 15 hours after waking, sees dopamine, norepinephrine and cortisol taper while serotonin rises, favoring relaxation; elevated cortisol late in the day is associated with depression and anxiety. Supportive practices include tapering bright artificial light and lowering lights in the room, viewing afternoon sun, NSDR such as yoga nidra, self-hypnosis through the free Reveri app developed by Stanford colleagues, heat like saunas or hot baths, and for some people ashwagandha, which lowers cortisol and which he cautions against using for more than two-week stretches without breaks. This phase suits habits that require less override, such as journaling or practicing music or a language you already do. People who prefer to exercise in phase two can, but he recommends some NSDR within an hour or two afterward.

Phase three, about 16 to 24 hours after waking, is when the rewiring happens during deep sleep. Keep the room dark and cool (the body drops one to three degrees to fall and stay asleep), don't go to bed hungry, avoid caffeine late, and consider sleep supports like magnesium threonate or bisglycinate, theanine and apigenin, which are detailed in the free sleep toolkit on his website. If you wake at night, use minimal light, since light quickly suppresses melatonin, and try a Reveri sleep script or NSDR to fall back asleep.

Once a habit feels effortless, Huberman says it's actually good to move it around the day. Being able to do it in any phase shows context independence. He explains that the hippocampus forms and holds procedures while they're being learned, and then the pattern of activity is transferred to the neocortex, at which point the habit no longer needs its bracket.

Reward Prediction Error: Rewarding the Whole Envelope

Huberman explains reward prediction error: an expected reward that arrives reinforces behavior, an unexpected reward produces an even bigger dopamine release, and an expected reward that fails to come drops dopamine below baseline. His example: if he's 50 minutes into a writing session and breaks his own rule to check social media, his dopamine falls below where it started. Much of the dopamine comes during anticipation, like a child excited about going to an amusement park.

To use this for habits, he suggests positively anticipating not just the habit but the time before and after it. For zone 2 cardio, which he likes less than weights, he acknowledges the start will be hard but focuses on how good he'll feel afterward, including enjoying his appetite. This is not positive self-talk: "you can't lie to yourself," he says, and lying sets up the opposite of reward. You have to be honest that the start is unpleasant while rewarding the whole sequence. He recommends writing out the steps before, during and after the habit, and treating the whole thing as the habit. Dopamine, he emphasizes, "is not about feeling good. It's about feeling motivated," and because epinephrine is made from dopamine, it also brings energy.

Program Two: The 21-Day System

The second program, used for a long time by someone Huberman knows well and their kids, involves choosing six new habits to do every day for 21 days, with the expectation that you'll complete only four or five each day. The goal is really the habit of performing habits. It also allows for things like resistance training that shouldn't be done daily. If you miss a day, there's no punishment, and you shouldn't try habit slip compensation by doing extra the next day.

He recalls advice from Berkeley neurologist Bob Knight when he started graduate school: don't adopt any routine you can't keep up consistently at least five and ideally six days a week. He also suggests thinking of the 21 days as a series of two-day chunks: if you can do it for a day, you can do it for two, then reset.

The unique part is what comes next: in the following 21 days you add nothing new and simply see which habits you keep doing automatically. Many people cram in too many new behaviors; this test shows whether you truly have room for more. Only once all six are reflexive would you add new habits. He suggests combining this with the three-phase plan.

Breaking Bad Habits

Foundational practices like stress reduction, good sleep and nutrition help, but Huberman wants a specific protocol. The mirror image of long-term potentiation is long-term depression (unrelated to mood): if neuron A fires and neuron B does not fire within a certain time window, their connection weakens. The challenge is that bad habits happen too quickly to catch in advance, like picking up your phone mid-work, often without noticing the steps.

He cites a 2020 scoping review by Heather Fritz, titled "intervention to modify habits," which found that reminders and notifications work in the short term but not over time. Punishments like shocks or fines work, but people rarely monitor and punish themselves consistently enough. What does work, he explains, is using the moment right after you catch yourself doing the bad habit to immediately perform a different, positive, easy behavior, such as drinking a glass of water, doing breath work or spending five minutes on a language. It seems counterintuitive, but it turns a closed loop into an open one and changes the whole sequence of firing neurons, like adding a new chord sequence each time you slip into the wrong one on the piano. Over time the bad habit takes on a new form or loses its pull. For addictions, he points listeners to his episode with Dr. Anna Lembke, noting that addiction requires a much fuller approach.

Conclusion

Huberman's framework treats habits as neuroplasticity that can be engineered. Recognize how much limbic friction a habit requires, lean on enjoyable linchpin habits, rehearse the steps once, and anchor new habits to phases of the day rather than exact times, putting the hardest ones in the alert first eight hours and easier ones later, then protecting sleep so the changes consolidate. Use reward prediction error honestly by anticipating the whole before-during-after envelope, try six habits for 21 days expecting four or five, then test for 21 days before adding more. And to break a bad habit, don't just scold yourself: follow it immediately with a small positive behavior to rewrite the circuit.

This is a summary of "The Science of Making & Breaking Habits" from Huberman Lab, giving you the GIST so you can decide if you want the REST. Listen to the full episode.