Why Does My Mind Wander So Much When I Meditate?
You sit down and close your eyes. Somewhere around the twentieth second you are thinking about an unanswered message, then about lunch, then about whether you are doing this correctly. You notice, you feel a small flash of irritation at yourself, and you start again. Four minutes later you have done that eleven times.
Most people read that as evidence they are bad at meditation. We think the reading is wrong, so we went and checked. If the drifting is a failure, our whole product is a five-minute failure machine.
Is it normal for your mind to wander?
It is more than normal. It is the resting state of a human being.
The 2010 Science study that put this question on the map sampled people's thoughts through an ordinary day using a smartphone. Its finding, in the authors' own words, is that people "are thinking about what is not happening almost as often as they are thinking about what is", and that doing so "typically makes them unhappy". That is normal waking life. Nobody in that study was meditating.
Now the number that ought to be printed inside every meditation app. In a 2026 study, 30 experienced meditators each sat two 30-minute breath-focused sessions and pressed a button every time they caught their own mind wandering. These were not beginners. They had completed a formal mindfulness programme at least two years earlier and reported 6.03 years of practice on average.
The average number of presses was 24.17 when sitting alone and 25.37 when sitting in a group. The difference between the two settings was not significant.
Twenty four catches in thirty minutes. That is roughly one every 75 seconds, which is our arithmetic rather than the authors'. In people with six years of practice behind them.
Read the spread as well, because it is the honest part. The standard deviation was about 19 presses. Some of those experienced meditators barely drifted at all. Others drifted constantly. Same programme, same years.
Does that mean the meditation failed?
No. This is where the research becomes genuinely useful.
A 2012 fMRI study at Emory put 14 practitioners in a scanner to meditate on the breath, pressing a button whenever they realised their mind had wandered. The researchers then cut the time around each press into four intervals: mind wandering, awareness of mind wandering, shifting of attention, and sustained attention.
Each interval showed something different. Mind wandering came with activity in default mode regions, the network linked to self-generated thought. The moment of noticing engaged salience network regions. The shift and the return engaged elements of the executive network. Four distinct phases, one loop, running over and over.
So "I keep drifting and catching myself" is not a broken meditation. It is the described structure of the practice. The button press is not an interruption of the exercise. It is the exercise.
One more detail from that paper: activity across those phases was modulated by lifetime meditation experience. The cycle is trainable.
Does it get better with practice?
Somewhat, slowly, and less dramatically than you would hope.
A 2018 study compared moderately experienced meditators with advanced ones across about an hour of seated concentration meditation, sampling their mental state with probes throughout. The expert group reported greater depth and frequency of sustained meditation. The non-expert group reported greater depth and frequency of mind wandering. The authors describe it as one of the first direct behavioural indices of meditation expertise.
So the thing does move. But notice what the comparison was: not beginners against course graduates, but moderately experienced practitioners against advanced ones.
At the short-training end, a 2023 analysis pooled five studies of a four-week mindfulness course run in a variety of organisational settings, 304 people in total. Compared with people who did no training, the trained group differed on self-reported mind wandering by a standardised mean change of -0.387, on meta-awareness by -0.374, and on the variability of their response times in an attention task by -0.376.
Here is the part a headline would skip. The trained group did not mostly get better. The untrained group got worse. In the authors' words, the no-training groups "showed increases in mind wandering and reductions in meta-awareness over time", while the trained participants stayed stable. The paper describes this as protection, not enhancement.
Two caveats, both raised by the authors themselves. These were not randomised controlled trials, and they call for such trials as the obvious next step. And two of the authors are the co-developers and copyright holders of the training programme being tested. They disclosed it. So are we.
Does getting annoyed at yourself make it worse?
This is the most practically useful thing we found, and it is the one nobody puts on a landing page.
A 2017 experiment at Carnegie Mellon randomly assigned 147 healthy young adults to one of four three-day trainings: mindfulness teaching both attention monitoring and acceptance, mindfulness teaching attention monitoring only, relaxation training, or reading as an active control. On the third day everyone completed a six-minute sustained attention task that measures mind wandering.
The group trained in monitoring plus acceptance had the lowest mind wandering, significantly lower than the group trained in monitoring alone.
Sit with that comparison. Both mindfulness groups learned to watch their own attention. Only one of them also learned to be accepting about what they saw. Acceptance was not the soft garnish on top of the real technique. It was the ingredient that moved the result. The authors suggest acceptance helps people manage the frustration of a boring task, which is a fair description of minute three of a five-minute sit.
The practical reading: the flash of self-criticism when you catch yourself drifting is not neutral. It is the part most likely to be working against you.
What does this mean for five minutes?
Take the expert rate and shrink it to the size of a Session. If a person with six years of practice catches their mind wandering roughly every 75 seconds, five minutes holds about four catches. Four repetitions of notice, accept, return. That is our arithmetic again, stretched from one study, so hold it loosely.
Still, it reframes what you are doing. You are not trying to hold an unbroken line of attention for five minutes. Nobody in this literature does that, including the people who have been at it for decades. You are running a handful of reps of a four-part cycle, and the wandering is the thing that makes each rep possible. A sit with no drifting in it would contain no repetitions at all.
It also means you cannot really fail one. The worst five minutes you have, the one where you were somewhere else the whole time and caught yourself eight times, contained twice the practice of a calm one.
What to do today
- Stop counting the drifts and start counting the catches. The catch is the event with its own brain signature, and it happens to six-year practitioners about every 75 seconds.
- When you notice, be uninterested about it. Not pleased, not annoyed. The one trial that isolated acceptance found that acceptance was the piece doing the work.
- Judge the session by whether you sat, not by how quiet your mind was. Quietness is not a variable anyone has shown you can command. Turning up is.
The evidence does not promise you a silent mind. It says a wandering mind is the ordinary state, that catching it is a distinct and trainable act, and that how you treat yourself in that half second matters more than how often it happens. Five minutes is room enough for about four of those. That is a practice, not a failed attempt at a longer one.
Sources
This article shares general information about meditation research. It is not medical advice. If you have concerns about your mental health, talk to a qualified professional.
Five minutes. Every day. That is the whole app.