Why there is no number on this page
You will find hour figures for various ages in circulation, and they are shorthand for real guidance that is more conditional than the shorthand suggests. This page does not state one as fact, for three reasons.
First, screen time is a poor measurement. A category that contains video calls with family, homework, a drawing app, a co-played game and an infinite feed is not measuring one thing, and studies that use it are averaging across activities with genuinely different effects. Second, most of the evidence is observational and cross-sectional. Families in which children use screens heavily differ from other families in income, stress, parental working hours and neighbourhood safety, and those differences also predict the outcomes being measured. Statistical adjustment helps and does not settle it. Third, when the same large datasets are analysed with different defensible choices, the results move — several well-publicised effects have shrunk considerably under reanalysis.
None of that means screens are fine and the question is closed. It means the honest position is uncertainty with some clear signals inside it, and that a parent making decisions in that condition is better served by a mechanism than by an hour count.
Displacement is the useful frame
The mechanism with the most support behind it is displacement. A screen hour is not harmful in itself so much as it is an hour that was spent on something else, and for children the something else is usually one of three things with substantial evidence behind them: sleep, physical movement, and back-and-forth interaction with people.
| What is displaced | Why it matters | What to look at |
|---|---|---|
| Sleep | The most consistent association in the literature. Devices in the bedroom, use close to bedtime and notifications overnight all track with less sleep and later sleep, and short sleep in children has well-documented consequences for mood, attention and learning | Where the devices spend the night, and what happens in the last hour before bed |
| Movement | Screen use is overwhelmingly sedentary, and the trade against outdoor and active play is direct | Whether there is active time in the day at all, rather than the screen total |
| Face-to-face interaction | For young children especially, language develops through responsive back-and-forth. A screen does not respond, which is the leading explanation for why the youngest children learn poorly from video compared with a person | How much of the day involves conversation, joint attention and being answered |
| Unstructured boredom | Harder to study and frequently mentioned by teachers and clinicians. Filling every gap removes the conditions in which children generate their own play | Whether there is any part of the day with nothing scheduled and nothing on |
| Homework and reading | Direct competition for the same hours, and the device usually wins by design | Whether the phone is in the room during homework |
The practical value of this frame is that it converts an unanswerable question into an answerable one. Rather than asking whether ninety minutes is too much, you ask whether this child is sleeping enough, moving during the day, and being talked with. If all three are intact, the residual worry about the total is smaller than it feels. If one of them is being eaten, that is the thing to fix, and it usually points at a specific window of the day rather than at the device in general.
For the youngest children
The one area where guidance is more definite involves infants and toddlers, and the reason is the video deficit — the well-replicated finding that very young children learn a word or a task considerably less well from a screen than from a person doing the same thing in front of them, and that the gap narrows when the screen is interactive and responsive, such as a live video call with a relative who answers.
Pediatric guidance for this age range therefore leans toward minimising passive screen use and treating video chat differently, and toward watching alongside the child rather than leaving them with it, so that an adult is doing the naming and answering that the screen cannot. That is a statement about how language develops, not a moral position, and your pediatrician is the person to ask about your child, particularly if there are any concerns about speech or hearing.
Realistically, screens get used in waiting rooms, on flights and during the hour when dinner has to happen. The distinction that matters is between occasional use and the device becoming the default way a child is settled, because once it is the only thing that works, the other ways stop being practised.
Content and co-use
Where research compares types of use rather than totals, the type usually matters more. Educational programming designed with child development input, video calls with family, a game played with a sibling and a drawing app are not equivalent to autoplaying algorithmic video, and treating them as the same category is the flaw at the centre of the hour-count approach.
Co-use is the other lever with support behind it. Watching or playing alongside a child converts a one-way stream into a conversation — you ask what happened, they explain, you say what you think about it — and that is the interaction the screen was otherwise displacing. It also means you know what they are watching, which becomes the entire game later on.
| Design feature | What it is doing | Practical response |
|---|---|---|
| Autoplay and endless feeds | Removes the natural stopping point that the end of an episode used to provide | Turn autoplay off where the setting exists. Choose a thing to watch rather than opening a feed |
| Streaks, dailies and limited-time events | Manufactures obligation, so stopping feels like losing something | Worth naming out loud with an older child. They usually recognise it immediately |
| Notifications | Fragments attention and pulls the device back into the room | Off during homework, meals and overnight. This is a bigger win than a time limit |
| Recommendation engines | Optimised for watch time, not for what you would choose for a nine-year-old | Assume drift toward more extreme content over a session, and check what actually got watched |
| In-app purchases and loot mechanics | Gambling-shaped reward structures aimed at children | Payment methods off the device, purchase confirmation required |
| Parental control tools | Useful as scaffolding, easily circumvented by a motivated teenager | Use them, and do not treat them as the plan |
Bedrooms, meals and the two rules that hold
If a household adopts only two rules, the two with the best evidence-to-friction ratio are that devices charge outside the bedroom overnight, and that meals are device-free for everybody including the adults.
The first protects sleep, which is the clearest association in the whole literature, and it also removes overnight notifications and the three-in-the-morning group chat. The second protects the conversational time that screens most directly displace, and it works only if it applies to the parents, because a rule that exempts the people making it is not a rule, it is a grievance.
The adult side of this is worth stating plainly. Children calibrate what devices are for by watching how the adults around them use them, and phone use during conversation is noticed and copied. If the household's screen problem is genuinely the children, that is worth checking honestly before the rules get written.
The first phone, and teenagers
There is no established right age for a first phone and the decision is driven mostly by logistics — how a child gets home, whether they are alone after school, what everyone else in their class has. Two distinctions are worth keeping separate: a device that makes calls and messages is a different decision from an account on a platform built to maximise engagement, and the second one is where most of the difficulty in the research lives.
For teenagers the evidence is contested in a specific way. Population-level associations between heavy social media use and worse mental health are real but modest on average, and they vary substantially between individuals — the same amount of use appears to affect different teenagers very differently, with the picture worse for those already vulnerable and for experiences involving comparison, exclusion or harassment rather than time as such. That variation is why an average is a bad guide to your own child, and why what they are doing and how they feel afterwards is a better question than how long.
The version of this conversation that works tends to be the one that treats the design as the adversary rather than the teenager: these products are engineered by large teams to be difficult to put down, which is a thing that is being done to them rather than a failure of their character. That framing keeps you on the same side, and it is the only position from which you will keep hearing about what actually happens online. If any of this becomes a question of mood, withdrawal, sleep collapse or safety, it stops being a media question and becomes a clinical one — a pediatrician or mental health clinician is the person for it.
What to do with a page like this
Write down what your household actually does for a week before changing anything, adults included, because nearly everyone estimates their own use badly and the specific hour that is causing the trouble is usually visible immediately. Then change one window rather than the whole picture: the hour before bed, or the meal, or the homework block. Building the plan with your pediatrician at a routine visit is the version of this that accounts for your child rather than for an average child.
The thing worth holding onto is that no study can tell you whether your ten-year-old is doing well. You can see that from across a room, and it is more information than any hour count on any page.
Questions people ask
How many hours a day is too much?
There is no threshold this page will state as fact, because the research does not support a clean one and because screen time as a measure lumps together activities with different effects — a video call with a grandparent and an autoplaying feed are not the same thing. The more useful question is displacement: is this child sleeping enough, moving during the day, and having real back-and-forth conversation. If those three are intact, the total matters less than it feels like it does. If one is being eaten, that is where to intervene.
Is the research on screens and children as clear as the headlines suggest?
No. Most of it is observational, so households with heavy use differ from other households in ways — income, stress, working hours, neighbourhood — that independently predict the outcomes being measured. Effect sizes are frequently small, and several widely reported findings have shrunk when the same datasets were reanalysed with different defensible choices. There are clearer signals inside the noise, particularly around sleep and around very young children learning less from video than from people, but the confident version of this story is not what the evidence supports.
Does it matter what they are watching, or just how long?
Where studies compare types of use rather than totals, the type usually matters more. Programming designed with child development input, video calls with family, a game played with a sibling and a drawing app are not equivalent to an algorithmic feed optimised for watch time. Co-use is the other lever with support behind it: watching or playing alongside a child turns a one-way stream into a conversation, which is exactly the interaction the screen would otherwise have displaced.
What are the highest-value rules for a household?
Two, and they are boring. Devices charge outside bedrooms overnight, which protects sleep and removes overnight notifications. And meals are device-free for everyone, adults included, which protects the conversational time screens most directly replace. The second only works if it binds the people who wrote it — children calibrate what devices are for by watching how adults use them, and a rule that exempts the adults reads as a grievance rather than a rule.
When is screen use a clinical question rather than a parenting one?
When it stops being about hours. Sleep collapsing, withdrawal from things the child used to enjoy, a persistent change in mood, falling function at school, or anything involving harassment, exploitation or safety online moves this out of the media-rules category. So does a young child about whom you have language or development concerns, since that needs an assessment rather than a screen limit. A pediatrician or mental health clinician is the right person, and a week of written notes about what actually happens is more useful to them than a recollection.