The Digital Whale

Screens and young people

How much screen time is too much

No dose-response threshold has been established for any age group, and the organisations behind the one and two-hour rules describe them as judgement calls.

There is no evidence-derived limit on screen time for any age group. The numbers in circulation — an hour a day for pre-schoolers, two hours for older children, none at all under two — come from guidance documents that describe them as precautionary judgements rather than measured inflection points. The World Health Organization grades the evidence behind its own under-fives recommendation as very low quality. The Royal College of Paediatrics and Child Health declined to give a number at all. A 2026 UK advisory panel published numbers and said in the same report that there is no convincing evidence for a threshold. Guidance has converged instead on displacement mattering more than duration. For a workplace-oriented comparison point on how digital activity is translated into metrics, see books about time management.

One study produced numbers from data, and they are fragile

The only widely cited thresholds estimated from a dataset came from a curve fitted to English 15-year-olds. The shape was an inverted U: a little screen use tracked slightly higher well-being than none, a lot tracked lower.

2017

Well-being rose then fell with screen use, with inflection points at 1 h 57 m for weekday smartphone use, 1 h 40 m for video games, 4 h 17 m for computers and 3 h 41 m for television, though screen use explained 1.0% or less of the variance.

Direction: Decrease. Strength of evidence: Mixed.

Przybylski & Weinstein, Psychological Science, 2017120,115 English 15-year-olds

Caveat Cross-sectional and self-reported; the inflection points are curve-fitting artefacts of one dataset and have not been consistently replicated elsewhere.

Two features are usually lost in quotation. The thresholds differ by device by more than a factor of two, which is not what a general limit would look like, and the variance explained is too small for any threshold drawn from it to carry weight for one child.

WHO's own grading undercuts its one-hour limit

The most-quoted limit for young children carries an internal contradiction, which WHO publishes rather than hides.

2019

WHO recommends no sedentary screen time under age 2 and no more than one hour a day at ages 2–4, classifying each as a strong recommendation based on very low quality evidence.

Direction: Decrease. Strength of evidence: Mixed.

World Health Organization, Guidelines on Physical Activity, Sedentary Behaviour and Sleep for Under-5s, 2019GRADE evidence review supporting global guidelines for children under 5

Caveat The mismatch between "strong recommendation" and "very low quality evidence" is WHO's own GRADE assessment, making the one-hour figure a policy judgement rather than a measured threshold.

The UK's paediatricians drew the opposite conclusion, stating that the magnitude of impact on key health outcomes is small and that the associations do not imply a causative link (RCPCH, 2019; the only accessible copy was a third-party mirror). They named displacement of sleep, exercise and socialising as the mechanism, and gave no number.

Of the outcomes that could support a threshold, sleep has the most consistent literature. It also has the weakest designs.

2015

90% of studies reported a significant adverse association between screen time and sleep outcomes, but 99% relied on self-reported or parent-reported measures and all but the three smallest lacked an experimental design.

Direction: Increase. Strength of evidence: Strong.

Hale & Guan, Sleep Medicine Reviews, 2015Systematic review of 67 studies published 1999–2014

Caveat Vote-counting across observational studies inflates apparent consistency, publication bias was untested, and the review predates smartphone-era measurement.

Agreement across studies sharing one flaw is agreement about the flaw as much as about the world. The size of the sleep effect is covered in what devices do to sleep.

Guideline adherence is measurable, and rare

Whether the two-hour rule is well founded is separate from whether children meet it. In the largest US child cohort most do not, and adherence to the full package of recommendations tracks cognition better than any single component.

2018

36% met the two-hour recreational screen time recommendation, 50% met the sleep recommendation and 17% met the physical activity recommendation, and meeting more recommendations was associated with better global cognition.

Direction: Increase. Strength of evidence: Mixed.

Walsh et al., The Lancet Child & Adolescent Health, 2018Over 4,500 US children aged 9–10 in the ABCD study

Caveat Cross-sectional and observational, so children with stronger cognition may simply choose different activities; the full text was CAPTCHA-blocked, so figures come from the press release.

Variance explained is too small to hang a threshold on

The deeper problem is that the relationship is faint. If screen use accounts for a fraction of a per cent of the variation in well-being, even a correctly located threshold would move very little.

2019

Technology use accounted for at most 0.4% of the variance in adolescent well-being across three large national datasets.

Direction: Decrease. Strength of evidence: Strong.

Orben & Przybylski, Nature Human Behaviour, 2019355,358 adolescents, three national datasets

Caveat Aggregate variance explained can mask larger effects in vulnerable subgroups, which the specification-curve approach was not designed to detect.

That caveat matters: a small average effect is compatible with a large effect on a small group, and no current design isolates that group. The evidence on adolescent mental health is where the argument is playing out.

Guidance has shifted from duration to displacement

Recent official documents have largely abandoned the clock. The panel that produced the newest numbers framed them explicitly as precaution.

2026

The panel concluded that how screens are used matters more than total duration, that use is most problematic when it displaces sleep, physical activity, parent-child interaction and creative play, and that there is "no convincing evidence" for specific time thresholds.

Direction: No detectable effect. Strength of evidence: Mixed.

Early Years Screen Time Advisory Group, EYSTAG report, 2026Expert advisory panel commissioned by the DfE and DHSC

Caveat Expert-panel consensus rather than meta-analysis, and displacement is asserted as the most plausible pathway rather than quantified.

The American Academy of Pediatrics made the same move in 2026, redirecting attention from time limits to design features such as endless scrolling, to family relationships and to platform accountability. That is professional consensus and advocacy positioning rather than new evidence, and it belongs next to what parents do with rules.

The short version

  • The only thresholds ever estimated from data ranged from 1 h 40 m for video games to 4 h 17 m for computers, and came from one unreplicated cross-sectional English study (Przybylski & Weinstein, 2017).
  • WHO classifies its under-fives screen limits as strong recommendations based on very low quality evidence — its own assessment of its own guidance (WHO, 2019).
  • Sleep has the most consistent literature, with 90% of 67 reviewed studies reporting an adverse association and nearly all relying on self-report (Hale & Guan, 2015).
  • Only 36% of US children aged 9–10 met the two-hour recreational screen recommendation, so the rule describes a minority (Walsh et al., 2018).
  • Nobody has measured a threshold at which harm begins, and the guidance documents that publish numbers say so themselves.