Meditation is attention training. The best evidence supports moderate improvements in anxiety, depression and pain — real, worth having, and roughly comparable to other active treatments rather than superior to them. One randomised trial found mindfulness training non-inferior to a standard SSRI for anxiety disorders. It also has documented adverse effects, which most guides leave out.
What is meditation actually doing?
Training a skill: noticing where attention has gone, and returning it. That is the entire mechanical content, and everything else is scaffolding around it.
The most common misunderstanding is that the goal is a blank or peaceful mind, which makes almost everyone conclude they are failing. Wandering attention is not the failure state. It is the repetition. Each time you notice you have drifted and come back, that is one rep — the noticing is the practice, not the interruption of it.
What follows from repeated reps is a small increase in the gap between a stimulus and your response to it. Not the removal of anger, anxiety or craving. A slightly better chance of seeing the thing arrive before you are already acting on it. That is a modest description and it is closer to what actually happens than most of what is written about it.
What does the evidence actually show?
Moderate benefits in specific domains, and no demonstrated superiority over other active interventions. The headline results are better than sceptics assume and considerably weaker than the marketing.
The most-cited systematic review and meta-analysis, published in JAMA Internal Medicine, found moderate evidence of improvement in anxiety, depression and pain from mindfulness meditation programmes, and low or insufficient evidence for effects on positive mood, attention, substance use, sleep and weight. Critically, it found no evidence that meditation programmes were better than active controls such as exercise, medication or behavioural therapy.
That last finding is the honest headline and it almost never gets quoted. Meditation works about as well as the other things that work. It is a legitimate option, not a superior one.
More recent randomised work strengthens the clinical picture: a trial comparing mindfulness-based stress reduction with escitalopram for anxiety disorders found the structured programme non-inferior to the drug. That is a genuinely strong result and worth knowing if you are weighing options with a doctor. Note the comparison is an eight-week structured course with an instructor, not ten minutes on an app.
The NCCIH summary of effectiveness and safety is a reasonable, non-promotional overview.
How long do you actually need to sit?
Less than the tradition implies and more consistently than most people manage. Frequency beats duration, particularly in the first few months.
The clinical trials with the strongest results generally use structured eight-week programmes with substantial daily homework — often around 30–45 minutes, plus a class. That is the dose behind the good numbers, and it is worth being honest that most people do not do it.
For someone starting alone, ten minutes daily is a defensible target, for a reason that has nothing to do with the ten minutes: a session short enough that you cannot justify skipping it is the one that survives a bad week. Six days of ten minutes does more than one heroic hour on Sunday. The case for short, frequent practice is made in more detail here.
Expect the first clear effects somewhere between two and eight weeks, and expect them to be unglamorous — recovering from irritation a little faster, noticing a spiral one step earlier.
Which kind should you do?
For most people starting out, focused attention on the breath — because it is simple, portable and the one with the largest evidence base behind it. The differences between styles matter less than whether you do it.
| Style | What you do | Best suited to |
|---|---|---|
| Focused attention | Rest attention on the breath; return when it wanders | Beginners; scattered attention |
| Body scan | Move attention systematically through the body | Physical tension; poor body awareness; falling asleep is fine |
| Open monitoring | Notice whatever arises without selecting | Established practitioners |
| Loving-kindness | Deliberately generate goodwill toward self and others | Harsh self-criticism, isolation |
| Noting | Silently label experience: “thinking”, “hearing” | Strong rumination; very busy minds |
Loving-kindness is the one worth flagging as underused: a meta-analytic review found it produces measurable increases in positive emotion. It is also the style most people find embarrassing, which is not a good reason to skip it.
Can meditation harm you?
Yes, and the failure to say so is the largest honesty gap in the entire field.
A systematic review of adverse events in meditation practices and meditation-based therapies found harms are real and substantially underreported — most trials did not even monitor for them. The varieties of contemplative experience study documented a wide range of meditation-related difficulties in committed Western practitioners, including anxiety, fear, distorted perception, dissociation and impaired functioning — in experienced people, some of them teachers.
Higher-risk situations: intensive retreats, long daily sits, a history of trauma, psychosis or dissociative disorders, and periods of acute crisis. Body-focused practice can be destabilising for people whose bodies do not feel safe, which is common after trauma — trauma-informed instruction exists and is worth seeking.
The rule this site applies: if practice consistently increases distress, dissociation or dread over several weeks, that is a stop signal — not a purification stage, not resistance, not a sign you are close to something. Reduce it, change style, or stop, and get support.
What the evidence doesn’t support
- “Meditation rewires your brain.” Overstated. Structural imaging findings exist but many come from small, cross-sectional studies of long-term practitioners, where self-selection is uncontrolled. The claim has run far ahead of the data.
- “It beats medication.” Not supported as a general claim. Non-inferiority in one anxiety trial is not superiority, and it does not generalise across conditions. Do not stop prescribed medication on the strength of it.
- “It cures physical disease.” No good evidence supports meditation treating disease. Effects on the experience of pain and on distress are better supported than effects on pathology.
- “More is always better.” Not supported, and adverse events cluster at the intensive end.
- “It makes people kinder or wiser.” Weakly supported at best. Advanced practice does not reliably produce good behaviour, as the conduct record of several prominent teachers demonstrates.
A protocol that survives contact with real life
- Same time, same place, every day. Attach it to something you already do without deciding — after the kettle boils, before the shower. Decisions are what fail, not intentions.
- Ten minutes. Set a timer and stop when it goes. Do not extend it on good days for the first two months; that is how a practice becomes conditional on enthusiasm.
- Breath at the nostrils or the belly. Pick one and leave it. Changing the object is usually avoidance.
- When you notice you have drifted, that is the rep. Return without commentary. A session with two hundred returns is a good session, not a failed one.
- Log one word afterwards. “Restless”, “dull”, “fine”. Takes three seconds and gives you a record that outlasts your memory of how it has been going.
- Review at eight weeks. Not for bliss — for whether you recover from irritation faster, sleep slightly better, or catch a spiral earlier. If there is nothing after eight consistent weeks, change style or accept that it is not your intervention. That is a legitimate outcome.
If sitting still is the barrier rather than the discipline, the low-effort approaches are covered here, and attention outside practice matters at least as much as attention inside it.
Ten years in, what it did and didn’t do
I have sat most days for about a decade. The honest accounting is less impressive than I would have written at year two and more valuable than I would have believed at year one.
What it did not do: make me calm, make me kind, remove anxiety, or produce a stable sense of peace. I am roughly as anxious as I was. Anyone who met me on a bad day would not identify me as someone with a long practice.
What it did do is narrow and genuinely useful. The gap between something happening and me reacting is wider than it was. I notice a spiral perhaps one step earlier than I used to — not at the start, but before it has consumed an afternoon. Recovery is faster. That is the whole return, and it compounds, because the difference between a bad hour and a bad week is mostly a matter of how early you catch it.
The years I wasted were the ones spent trying to have better sessions. Session quality turned out to be almost entirely unrelated to benefit. The dull, distracted, clearly-failing sits appear to have done the same work as the good ones. I have never quite forgiven the practice for that, and it is the single most useful thing I know about it.
Frequently asked questions
I can’t stop my thoughts — am I doing it wrong?
No. Stopping thoughts is not the aim and is not achievable. The task is noticing that attention has wandered and returning it, which requires thoughts to wander in order to work at all. A busy session is a session with many repetitions. The only way to fail is not to sit.
Does it matter what time of day?
Only in that morning practice is more likely to actually happen — the day has fewer opportunities to take it from you. Beyond that, consistency matters more than timing. Late-evening practice makes some people alert rather than sleepy, so if you are using it for sleep and it is not working, try moving it earlier.
Are apps worth it?
They are a reasonable way to start, and the reminders and streaks genuinely help with the hard part, which is showing up. The limitation is that the strong clinical results come from structured eight-week courses with a live instructor who can see when something is going wrong. An app cannot notice that you are dissociating.
What if it makes me anxious?
Common early on, and usually it settles — sitting still removes the distractions that were covering background anxiety, so you meet it. If it persists or worsens over several weeks, treat that as a signal rather than an obstacle. Try shorter sessions, eyes open, a body scan or walking practice, and consider trauma-informed instruction.
Is spiritual meditation different from mindfulness?
The techniques overlap almost completely; the framing differs. Secular mindfulness strips the metaphysical context to make the practice portable and researchable. Traditional forms embed it in a system with ethical commitments and a goal beyond stress reduction. The mechanics of sitting are largely the same. What differs is what you are told it is for.
Important: do not stop or reduce prescribed medication on the basis of anything here. Meditation has documented adverse effects and is not appropriate as a sole approach for trauma, psychosis or acute crisis. In crisis, contact a helpline in your country. See the editorial policy.
