A commuter downloads a meditation app during a stressful week, follows a ten-minute session on the train for three mornings running, and then starts swiping the reminders away. Whether those three mornings did any good, and whether the subscription is worth renewing, are fair questions.

Do meditation apps work, then? Yes, modestly: for everyday stress, worry and low mood, over a few weeks, and only while you keep using them. A 2024 meta-analysismeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary of 45 randomized trials, led by Jake Linardon at Deakin University in Australia, found that mindfulness apps reduced symptoms of depression and anxiety slightly compared with control groups, mostly in people without a diagnosis. Nobody yet knows whether the benefit lasts beyond about two months, and apps have not been shown to match or beat other therapies.¹

The practical reading is short. Decide when you will practice before you pay, give the app a few weeks of regular use, and keep it only if you are still opening it and feel steadier. The checklist further down turns that into six quick tests.

Where do meditation apps work, and by how much?

The clearest evidence that meditation apps work comes from people without a diagnosis, and the effect is small: in the 2024 Deakin University meta-analysis, people given a mindfulness app reported slightly fewer symptoms of depression and anxiety than comparison groups, measured after 10 days to 8 weeks. The result held in the larger and better-designed trials, but too few trials followed people up afterwards to say whether it lasts.¹

The study

Moderate evidence

In 45 trials, one extra person improved for every 11 to 14 app users

Across trials with 5,852 people for depression and 6,082 for anxiety, app users improved more than control groups by a small margin (effect sizes of 0.24 and 0.28). The authors translate this into numbers needed to treat of about 14 and 11: more than 10 people would need to use an app for one extra person to improve compared with a control group. Results held in trials at lower risk of bias and in larger trials. Against another therapy, tested in only three or four small trials, there was no clear difference either way.¹

In plain terms, most people who use an app would have felt much the same without it, and a minority gets a lift they would not otherwise have had; nobody can tell in advance who. The Deakin team notes that apps built on cognitive behavioral therapy show effects of a similar size, so a small benefit is typical of self-help on a phone rather than a flaw peculiar to meditation. The authors also point out what most apps leave out compared with a face-to-face course: they are self-guided, with no teacher or group to talk to, and they lack the cognitive-therapy elements that courses such as MBCT include.¹

Face-to-face courses remain the benchmark, and what mindfulness is and what its trials found covers how they compare with other treatments.

Two other reviews agree on the direction. A 2021 meta-analysis of 34 trials, led by Eva Gál at Babeș-Bolyai University in Romania, found the largest effect on perceived stress and no clear effect on distress or general well-being; we had access only to its abstract.² A 2024 review of trials in college students judged the certainty of evidence moderate for stress, depression and well-being.³

Picture a team of 12 who all start an app through a workplace perk and use it as trial participants did. On the Deakin figures, roughly one of them would improve who otherwise would not, and nobody could say in advance which one. So judge an app by what it does for you: after a few weeks of regular use, do you feel steadier than before?

A waitlist is an easy opponent

How good a meditation app looks depends on what it is tested against. In the 2024 Deakin review, about half the trials compared app users with people on a waitlist, and only four compared an app with another therapy, finding no clear difference either way.¹

A waitlist controls for very little. People handed an app also get attention from researchers, a daily routine, a reason to pause and the expectation of feeling better, and some researchers argue that part of an app’s effect comes from using the device itself, a “digital placebo”. In a secondary analysis, trials with an active comparison groupcontrol group: The group in a study that does not get the treatment being tested, so its results show what would have happened anyway. It may get nothing, usual care, a placebo, a place on a waiting list or a different activity, and which one it gets shapes what the comparison can show.Full entry in the glossary showed somewhat smaller effects than waitlist trials, though the difference was not statistically clear.¹

1. Against a waitlist: people who wait get nothing to do, the easiest test for an app; about half the trials in a 2024 review of 45 app trials used this comparison
2. Against a matched activity: a similar activity such as relaxation audio; in a 2019 trial by UK researchers, a muscle-relaxation recording eased anxiety and depression about as much as a mindfulness app
3. Against another therapy: only four small trials in the 2024 review made this test, and they found no clear difference either way

A trial by two UK researchers shows what a fairer test looks like. In 2019, Cristina Quinones (Open University) and Mark Griffiths (Nottingham Trent University) randomly assigned 994 adults who worked long hours and showed early signs of compulsive internet use to a mindfulness app, a muscle-relaxation recording or a waitlist, for 10 minutes a day over two weeks. Both practices eased anxiety and depression compared with waiting, and the app did no better than relaxation on those symptoms, though it did more to curb compulsive internet use. The Richard Benjamin Trust funded the study and the app’s maker gave participants free access. All measures were self-reported, and dropout was heavy: only 64 of the 343 people given the app answered the final questionnaire, a far smaller share than in the other two groups, which weakens the comparison.⁴

Marketing compresses all of this into phrases like “clinically proven”. A trial against a waitlist shows that an app beat doing nothing, not that it beat a free relaxation recording, so before trusting a research claim on an app’s website, find out what the comparison group got.

Myth

An app backed by research has been shown to beat other ways of relaxing.

Fact

About half the app trials in a 2024 review compared the app with a waitlist, so a research claim may only mean it beat doing nothing.

Most downloads go quiet within ten days

Most people stop opening mental health apps within about ten days of installing them, a 2019 analysis of real-world use by Amit Baumel and colleagues found. Across popular Android apps for anxiety, depression and well-being, the share of users opening an app fell by more than 80 percent between the first day and the tenth.⁵

Mindfulness apps kept users better than breathing-exercise apps, and their users spent longer in them each day, yet only about 1 in 20 was still opening one a month after installing it. The figures come from a commercial panel of Android users worldwide in 2017 and 2018, which could not show who the users were or why they stopped.⁵

Trials lose people too. A 2023 review of 70 randomized trials by Linardon found that about a quarter of those given an app dropped out, more in larger trials and in samples drawn from the general adult population, and fewer in trials that paid participants or enrolled them in person. Engagement was poorly reported, but the review found clear problems with keeping people using the apps; only the abstract of this review was available to us.⁶

Why such a steep drop? Installing takes seconds, so a download says little about whether someone has the time, motivation or a free slot in the day to practice, Baumel’s team suggests.⁵ On these figures, the commuter at the start of this article is the typical user, not the exception.

An app can only help on the days you open it, so plan the habit before the purchase. Tie the session to a fixed point you already pass each day, such as the train ride or the first coffee, the same logic as building any habit through repetition, and keep sessions short; the few trials that compared lengths found little advantage in meditating for longer. Decide when you will practice before you decide what to pay.

Decide when you will practice before you decide what to pay.

One app dominates the research, often with its maker’s help

Much of the evidence on meditation apps comes from trials of one commercial app, and many of those trials had ties to its maker. A 2022 systematic review led by Alison O’Daffer at Seattle Children’s Research Institute in the US found 14 randomized trials of Headspace and only 1 of Calm, the two market leaders at the time; how those two compare with each other is left to a separate comparison.⁷

Half of the Headspace trials declared a conflict of interest involving the company, most often free premium access for participants. In two, company staff helped design, run or analyze the study and were listed as co-authors, although the reviewers found no trial clearly funded by the company. Only about a third of the trials were preregisteredpreregistration: Filing a study's hypotheses and analysis plan publicly before the data are collected or examined. It stops a planned test from being rewritten once the results are in, so a reader can tell a real confirmation from a hunt through the data.Full entry in the glossary, and most were too small to detect small or medium effects.⁷

Free access shapes the evidence even without money changing hands. Researchers are more likely to study an app they can get for free, the reviewers argue, so one app piles up trials while others, possibly just as good, go untested. A long list of studies can make an app look like the best-supported choice when it is mainly the most-studied one.⁷

Two findings keep this in proportion. In a secondary analysis in the Deakin review, trials of the most-studied app produced effects much like trials of other apps, and the same authors note that most mindfulness apps on sale have no controlled-trial evidence behind the benefits they advertise.¹

So when an app’s website lists dozens of studies, count how many tested that app rather than mindfulness in general, and check whether the company helped run them.

Checking a meditation app before you subscribe

The questions researchers ask of a trial work for a shopper too. These are suggestions for healthy adults trying an app for everyday stress, drawn from the evidence above; they are not a way to treat a mental health condition.

WiserHours

Six checks before subscribing

• Look for a randomized trial of this app itself, not of mindfulness in general
• Check what the comparison group got: a waitlist, or something active such as relaxation audio
• Check who ran and paid for the trial, and whether it was preregistered
• Pick a fixed daily slot, such as the commute or the first coffee
• After a few weeks, keep it only if you still use it and feel steadier
• Use the app alongside any treatment you have, not instead of it

Find this checklist online:
https://wiserhours.com/mindfulness/meditation-apps/#checklist-meditation-apps-before-you-pay

How the claims you are likely to meet in app marketing hold up:

Duration, diagnosis and harm: three open questions

The trials leave three large gaps: how long any benefit lasts, whether apps help people with a diagnosed condition, and how often they do harm. The 2024 Deakin review could measure effects only at the end of each program because so few trials followed people up; most enrolled students or members of the public, and very few reported how many people recovered or got worse.¹

Trial conditions also differ from everyday use. In most trials of the leading app, participants got the paid version and were told which sessions to follow, whereas at home you choose freely.⁷

Harms are rarely counted. A 2024 review of 171 trials of mental health apps of all kinds found that only 55 reported adverse events at all.⁸ The UK’s NHS warns that some people find mindfulness unhelpful, and some come away feeling worse.⁹ Treat an app as a short-term aid whose long-run value is unknown, and if sessions start to unsettle you rather than settle you, pause them and raise it with a doctor.

Signs you need a person, not a program

A meditation app is a self-help tool, not a treatment, and some situations call for a person rather than a program. Sorted by urgency:

• Right away: if thoughts of suicide come up or you fear you cannot stay safe, contact your local emergency number now, or go to an emergency department. As a US example, the National Institute of Mental Health lists 911 for emergencies and the 988 Suicide & Crisis Lifeline (by phone or text).¹⁰ As a UK example, NHS advice names 999 and A&E for emergencies, and the Samaritans listening line, 116 123.¹¹ Anywhere else, a doctor or local health service can name the crisis lines that serve your area.
• Within days or weeks: if sadness has hung on for weeks rather than a few days, or things you once enjoyed have lost their pull, book a doctor’s appointment; the UK’s NHS advises against putting it off.¹²
• Next time you see a clinician: if you already have a mental health condition or are in treatment, mention the app and use it alongside care, not in place of it. The Deakin authors conclude that apps alone are unlikely to meet mental health needs across the population and that other treatments are still required.¹

The bottom line

Meditation apps can help a little, for a few weeks, if you keep using them, which most people do not. The evidence is clearest against doing nothing and thin against other options, and a large share of it comes from trials of one app, often with ties to its maker. Treat an app as a low-cost practice tool, decide when you will use it before you pay, and take low mood that lasts to a doctor rather than an app.