Do Digital Detoxes Actually Work?

Mostly not, on the existing research, and the one intervention that produced large effects was narrower and better controlled than anything called a detox. A systematic review of 21 studies found that most showed either no effects or mixed findings on detox efficacy, and that this held equally across randomised and non-randomised trials. A 2025 randomised controlled trial that blocked mobile internet for two weeks found substantial improvements in mental health, wellbeing and measured attention.

So the honest answer is that “digital detox” describes a category too loose to evaluate, and that the specific, tightly defined version of it worked.

Why most of this research went nowhere

The 2022 systematic literature review by Radtke and colleagues in Mobile Media and Communication gathered the digital detox intervention studies and reported results that were, in the authors’ word, diverse.

Their summary is worth having in full. Even though a few more studies revealed positive rather than negative consequences from digital detox interventions, most of the studies showed either no effects or mixed findings regarding detox efficacy. Those diverse and contradictory findings were equally evident across randomised controlled trials and non-randomised controlled trials, as well as across studies with different sample sizes, similar intervention periods, and different subsets of smartphone use.

That last clause matters. When findings scatter regardless of study design, sample size, duration and what exactly was given up, the problem is unlikely to be any single methodological weakness.

The review covered 21 studies, of which 12 were randomised controlled trials, with the remaining nine non-randomised and all carrying pre- and post-intervention measurements. Study quality was formally assessed, and none fell below the tool’s 55 per cent cut-off, with six studies above 75 per cent. So this is not a case of pooling bad research.

The review also notes the definitional problem directly: it uses digital detox as an umbrella term covering abstinence, breaks, disconnection, timeouts and unplugging, and defines it using the dictionary sense of refraining from electronic devices to reduce stress or focus on the physical world. When a review has to build an umbrella that wide, the studies underneath are not testing the same thing.

What a more recent review found

A 2025 review searched three databases to August 2024 for digital detox methods and activities suitable for adults in a working population, outside the workplace. It included 12 studies covering 3,040 participants, spanning 2019 to 2024, and identified six categories of detox method, such as device restriction and app control, and seven categories of activity, such as physical and social activities.

Its conclusion on effectiveness: the heterogeneity of study designs and measured outcomes precluded a definitive evaluation of the effectiveness of these strategies.

Three years after the first review, with newer studies and a working-adult focus, the field still cannot say whether detoxes work. That is a meaningful update: the problem is not that the research is young.

The trial that found something

A 2025 randomised controlled trial in PNAS Nexus by Castelo and colleagues took a different approach. Rather than asking people to abstain, it used an app to block all mobile internet access on participants’ phones for two weeks, and tracked compliance objectively rather than by self-report. Texts, calls and internet on other devices were untouched, so the intervention isolated mobile internet specifically. It was preregistered, ran a month, and used a crossover design where a delayed group acted as control then received the intervention.

The results were substantial. The intervention improved mental health, subjective wellbeing, and objectively measured ability to sustain attention, with 91 per cent of participants improving on at least one. The attention measure was a task, not a questionnaire. Average screen time in the intervention group fell from 314 minutes a day to 161. Effects replicated when the delayed group received the intervention.

The authors put the sizes in context themselves: the change in measured sustained attention was about the magnitude of ten years of age-related decline, and the effect on depression symptoms was larger than the meta-analytic effect of antidepressants and similar to cognitive behavioural therapy.

Read those comparisons carefully

Those two comparisons are the most quotable sentences in the paper and the most likely to be misused, so here is what the authors say immediately afterwards: while the comparisons are thought-provoking, the nature of their sample and intervention is quite different from those studied in clinical psychology contexts.

That is the authors telling you not to do the thing the headline invites. Nobody should read this as evidence that blocking mobile internet substitutes for treatment. If you are being treated for depression or anxiety, or think you should be, that is a conversation with a doctor, and nothing in a two-week phone study changes it.

The paper is candid about other limits. Compliance was poor: of 467 people who committed to blocking mobile internet, 266 set up the app and 119 met the preregistered definition of compliant, with around 42 per cent never signing up at all. The sample was self-selected and motivated, with 83 per cent rating themselves five or higher out of seven on motivation to reduce phone use. Participants knew the study was about how smartphones affect wellbeing, and the authors name expectancy effects as a design limitation. Those who completed all three surveys had better baseline mental health and attention than those who dropped out, which the authors say limits generalisability.

Notably, the main effects held even including people who did not comply. The authors read this as suggesting that fully blocking mobile internet is not necessary to produce benefits, and that simply reducing use may be sufficient. That is a different question again from whether the phone on your desk costs you anything while you are not using it.

What seemed to be doing the work

The trial tested mechanisms, and the answer was mostly not about screens.

Blocking mobile internet increased time spent in what the researchers called the offline world, a factor covering socialising in person, exercising, being in nature, pursuing a hobby and reading. It decreased media consumption, increased social connectedness, increased feelings of self-control, and increased sleep. The effects on wellbeing and mental health were mediated by those changes.

Sustained attention was the exception. None of the measured mechanisms explained the attention effect, which the authors flag as suggesting different pathways for different outcomes.

That mediation result reframes the whole question, and it lines up with the reviews. What improved was not the absence of a phone. It was what filled the hours, which is precisely the variable the earlier review identified as uncontrolled across most detox studies. It also sits alongside after-hours messaging and burnout, where the pressure turned out to live in expectations rather than in the device.

Common questions

Is a weekend digital detox worth doing?

The evidence on short abstinence is genuinely mixed. A 2022 systematic review of 21 studies, 12 of them randomised controlled trials, found that most showed either no effects or mixed findings on detox efficacy, and that this held equally across randomised and non-randomised designs and across different sample sizes and intervention periods. A 2025 review of 12 studies covering 3,040 working adults concluded that heterogeneity of study designs and outcomes precluded a definitive evaluation of effectiveness. The clearest positive result comes from a two-week randomised trial blocking mobile internet, which is longer and more specific than a weekend off.

Do I have to block everything for it to work?

Apparently not. In the 2025 randomised trial, of 467 people who committed, 266 set up the blocking app and only 119 met the preregistered definition of compliant, with around 42 per cent never signing up at all. The main effects held anyway in the analysis including non-compliant participants, and the authors concluded that fully blocking mobile internet is not necessary to produce benefits and that simply reducing use may be sufficient. They also note that less restrictive approaches might be easier to sustain and could work as well, though that has not been tested.

Was it the lack of phone, or something else?

Mostly something else, on the trial’s own analysis. Blocking mobile internet increased time spent socialising in person, exercising, being in nature, pursuing hobbies and reading, decreased media consumption, and increased social connectedness, self-control and sleep. The effects on wellbeing and mental health ran through those changes rather than being direct. The effect on sustained attention was the exception, with none of the measured mechanisms explaining it. This matches the 2022 review’s own diagnosis, which identified the loose definition of detox as a problem: it covers abstinence, breaks, disconnection, timeouts and unplugging as one category.

The study said the effect beat antidepressants. Should I try this instead of treatment?

No, and the authors say so themselves. The paper reports the effect on depression symptoms was larger than the meta-analytic effect of antidepressants and similar to cognitive behavioural therapy, then immediately adds that while these comparisons are thought-provoking, the nature of their sample and intervention is quite different from those studied in clinical psychology contexts. The sample was self-selected and motivated, participants knew what the study was about, and the authors list expectancy effects as a limitation. If you are being treated for depression or anxiety, or think you might need to be, that is a conversation for a doctor. Nothing here is a reason to change or stop treatment.

The hours, not the device

The framing of a detox is abstinence: the phone as a substance, the fortnight as a cleanse. The research does not support that shape.

Two reviews, three years apart, could not establish whether detoxes work, and both located the problem in the same place. The 2022 review found results scattering equally across randomised and non-randomised designs, sample sizes and intervention periods, while using detox as an umbrella covering five different practices. The 2025 review found heterogeneity of designs and outcomes precluding any definitive evaluation.

Meanwhile the one trial that defined its intervention tightly, blocked one specific thing, and measured compliance objectively found people socialised, exercised, went outside, read, slept more and felt more in control, and those changes carried most of the benefit.

Which suggests the detox framing has it backwards. The question is not how long you can go without, and the answer is unlikely to be found by testing your willpower against an app. It is what the reclaimed hours are for. On the evidence available, a detox with a plan for the time is a different intervention from one without, and only the first has been shown to do much. If the underlying worry is how small the screen time effect actually is, that is worth reading before committing to a fortnight of anything.