Screen time before age two: what the evidence says

The question

Does screen time before age two help, harm, or do nothing to a child's development — and does it matter what kind of screen time it is?

Short answer

More passive screen time before age two is linked to slightly weaker language development — but the link is small, and nearly all the evidence is observational, so screens have not been proven to cause delays [1, 2]. What the evidence can't do is tell you whether a given amount of screen time is a net positive or negative for your individual child — only that heavier use is associated with slightly weaker language at population level. Toddlers learn almost nothing from passive video compared with learning from a live person [8], while interactive video chat with a real person can support word learning [5, 6]. The practical takeaway: keep solo passive viewing as low as you can, watch with your child when the screen is on, and don't leave the TV running in the background [1].

What the strongest evidence says

The overall association is small and negative. A meta-analysis of 42 studies with about 18,900 children found that more screen time was associated with slightly lower language skills (correlation r = −0.14), and that background TV — a television on while the child does something else — was also negatively associated with language (r = −0.19, though from only 5 studies) [1]. To put r = −0.14 in plain terms: screen time explains roughly 2% of the variation in children's language scores. That's the kind of association that shows up across a population but is invisible in any single child [1].

A large prospective study found a dose-response pattern. In Japan, 7,097 children whose screen time was reported at age 1 were followed to ages 2 and 4. The odds of communication delay at age 2 rose steadily across exposure groups (from under 1 hour a day up to 4+ hours), and communication and problem-solving delays were still elevated at age 4 for the higher-exposure groups [2]. But this is still observational: mothers of high-screen-time babies were also more likely to be younger, first-time parents with less education and lower incomes, and more likely to have postpartum depression [2] — all things that independently predict language outcomes. The study reported odds ratios only, not the absolute rates of delay in each exposure group, so there is no "X in 100" figure to quote. It also relied on parents to report both the screen time and the developmental outcomes, and recorded nothing about what was watched or whether a parent watched along [2].

The video deficit: toddlers don't learn from passive video like they learn from people. In a landmark experiment, 9-month-olds exposed to Mandarin through live speakers learned to distinguish Mandarin sounds, while babies exposed through video or audio alone learned nothing [8]. The pattern holds for word learning: 24- to 30-month-olds learned novel verbs from live interaction and from socially responsive video chat equally well, but showed no learning from non-interactive video of the same material [5].

Video chat is not the same as passive viewing. In a follow-up experiment with 17- to 25-month-olds, children who had a live FaceTime conversation with an adult recognized that adult and showed word and pattern learning a week later; children who watched a prerecorded video of the same speaker did not [6]. This is why the American Academy of Pediatrics carves video chatting out of its screen-time limits [11]. (The WHO guideline does not state an explicit video-chat exception [12].)

But video chat isn't magic either. In a larger experiment with 176 toddlers, adding social responsiveness to a video presentation (a closed-circuit, video-chat-like setup) was not enough to support word learning — children learned from a responsive live person, but not from responsive video [7]. So interactivity helps, but a screen still isn't a person.

Co-viewing and content quality change the picture. In the same meta-analysis, educational programming (r = 0.13) and watching together with a caregiver (r = 0.16) were associated with stronger language skills, and starting screens later was also associated with better outcomes (r = 0.17) [1]. A systematic review of 18 studies with over 32,000 children concluded that passive exposure was particularly harmful for vocabulary and comprehension in the first two years, while supervised, interactive exposure to age-appropriate content could be beneficial [4]. A separate meta-analysis of 28 studies in 0- to 3-year-olds found an overall negative association, but again found co-viewing and later start age were positive, and found no significant link for educational program viewing specifically [3].

The likely mechanism is displacement, not toxicity. The leading explanation researchers offer is that screen time replaces the activities that actually build language: face-to-face talk, play, and reading together [1, 10]. Observational studies can't separate a direct effect of screens from this displacement, from reverse causation (harder-to-engage children getting more screens), and from confounding — so the concern is as much about what screens crowd out as about what they do directly. One small brain-imaging study of preschoolers found more screen use was associated with differences in brain white-matter structure, but it was cross-sectional and can't show screens caused the difference [9].

Official guidance is conservative and mostly consensus-based. The WHO recommends no sedentary screen time for children under 1, and none for 1-year-olds either; at age 2, no more than an hour a day ("less is better") [12]. The AAP advises avoiding screen media other than video chatting before 18 months, and for 18–24 months only introducing high-quality programming watched together with a parent [11]. These are expert-consensus recommendations — sensible, but not the product of randomized trials.

What it means for the parents (not just the child)

The short version: the research says the parent's screen use and mood may matter at least as much as the child's — but almost none of it studies children under 2 specifically.

• Background TV quiets you, not just the room. In a lab study of 49 families, parents used fewer words — and fewer new words — per minute with background TV on [14]. In real homes, day-long recordings showed each hour of audible TV meant roughly 770 fewer words spoken by adults to the child — about a 7% drop from the usual rate [15]. It's not that the TV damages anything directly; it just displaces the talking that builds language. (One conflict to know: four authors of the recordings study were employed by the company whose recording devices were used [15].)
• Your own phone counts too. When researchers started studying "technoference" — a parent's device use getting in the way of parent-child interaction — the associations were consistent: parents absorbed in phones were less responsive and more likely to be harsh or snappy, especially during discipline [16][17]. But nearly all of this research is cross-sectional snapshots, so the arrow could run both ways — and underlying stress and depression may drive both the phone use and the parenting pattern [17]. Fathers who reported more technoference had toddlers with more behaviour problems (though harder children may drive parents to their phones) [16].
• Fussier babies get more screens — and the arrow goes both ways. In 744 caregiver-child pairs, more "difficult" child temperament at baseline predicted more handheld-device use later [19]. And in the best-known longitudinal study, heavier TV viewing at 6 months predicted more parental concern about the child's language a year later — but toddlers who were already struggling also got more TV time, so "screens hurt development" is too simple a reading [20].
• Your mental health may matter more than the screen minutes. In a 2025 systematic review, most studies of parental factors and children's problematic screen use found that parental stress, depression, and anxiety tracked with more problematic use in the child — though almost all the evidence was cross-sectional snapshots, not experiments [21]. (Parental-outcomes evidence rating: C.)

The bottom line for you: for under-2s, the most honest reading is that screens are mostly a delivery system for displaced interaction — yours as well as theirs. If the TV is on during play, you're talking less. If you're on your phone during the witching hour, you're more snappy. And if you're reaching for the tablet because you're running on empty, that's not a moral failure — it's the most common reason parents use screens, and it's the least studied one.

What remains uncertain

Causality. No randomized trial has assigned babies to different amounts of screen time and followed their development for years — and such a trial would be hard to run ethically and practically. Everything we know about long-term effects comes from observational studies, which means reverse causality is always on the table: a child who is harder to engage, or who shows early signs of language delay, may be put in front of screens more often, not the other way around [2, 4].

Which screens and which content. Most research studied television. Evidence on tablets, phones, and interactive apps in under-2s is thin, and the meta-analyses reach partly different conclusions about educational content [1, 3]. We also don't know how much of the effect is about the device versus what's on it.

How much is "too much." The dose-response in the Japanese study looks smooth — more hours, more risk [2] — but no study has identified a threshold below which screen time is definitely harmless or above which it is definitely harmful.

Absolute risks. The studies report odds ratios and correlations, not absolute rates of delay in each exposure group, so there is no reliable "X in 1,000" figure to give you. This is itself a gap in the literature worth knowing about.

Screens as parental respite. No study has tested what screens replace when parents are exhausted: there is no evidence on screens as deliberate parental respite — whether 20 minutes of cartoons that let a parent decompress is better or worse than a frayed, resentful alternative. (Evidence gap D.)

Benefits and risks in absolute terms

Honestly? We can't fill in this table the way the editorial rules demand, because the literature doesn't provide absolute risk numbers — only correlations and odds ratios. What can be said:

• Passive, solo viewing: associated with slightly lower language scores (about 2% of variance explained); in one large cohort, the odds of communication delay rose with daily dose [1, 2]. The studies report odds ratios and correlations only — not absolute delay rates per exposure group.
• Background TV: negatively associated with language skills, including by reducing how much and how well parents talk to toddlers [1].
• Video chat with a responsive adult: supported word learning in small experiments [5, 6]. No known developmental downside has been demonstrated.
• Co-viewed, high-quality content: small positive associations with language in observational data [1].
• Documented benefit of screens for under-2s: essentially none for passive viewing; the "educational video" products marketed to babies have no strong evidence behind them [4, 8].

Practical considerations

• Video calls with grandparents are fine. The AAP exempts video chatting from its screen-time limits (the WHO states no explicit exception), and experiments show toddlers can learn from responsive video interaction [5, 6, 11].
• If the screen is on, watch together and talk about it. Co-viewing is the one screen habit with a positive association in the data [1, 3].
• Turn off background TV. It's associated with poorer language outcomes and with parents talking less to their toddlers [1].
• Protect the things that build language: conversation, play, and reading together. One study of infants found that time with screens was ambiguous for vocabulary, while shared book reading was consistently positive — the opportunity cost runs one way [10].
• Don't buy "educational" baby videos expecting results. The evidence that they teach under-2s anything is weak to nonexistent [3, 4, 8].
• A short video to get through cooking dinner is not a proven crisis. The associations in the data are small and population-level; existing studies can't estimate the effect of modest, occasional viewing on an individual child. Parental sanity is a real input to good parenting.
• Mind your own phone during play, not just theirs. "Technoference" is consistently associated with worse interaction quality — but the evidence is mostly cross-sectional and bidirectional, so read it as a pattern to watch rather than a proven cause [16, 17, 18].

When to talk to your doctor, midwife, or pediatrician

Talk to your pediatrician if you're worried about your child's language development at any point — for example, if they aren't babbling, gesturing, or responding to their name toward the end of the first year, or aren't using words by around the middle of the second year [13]. Early evaluation is always better than waiting, whatever the cause.

References

1. Madigan S, McArthur BA, Anhorn C, Eirich R, Christakis DA. Associations between screen use and child language skills: a systematic review and meta-analysis. JAMA Pediatrics. 2020;174(7):665-675. doi:10.1001/jamapediatrics.2020.0327 — [B]
2. Takahashi I, Obara T, Ishikuro M, et al. Screen time at age 1 year and communication and problem-solving developmental delay at 2 and 4 years. JAMA Pediatrics. 2023;177(10):1039-1046. doi:10.1001/jamapediatrics.2023.3057 — [C]
3. Jing M, Ye T, Kirkorian HL, Mares M-L. Is screen exposure beneficial or detrimental to language development in infants and toddlers? A meta-analysis. Early Child Development and Care. 2024. doi:10.1080/03004430.2024.2349622 — [C]
4. Massaroni V, Delle Donne V, Marra C, Arcangeli V, Chieffo DPR. The relationship between language and technology: how screen time affects language development in early life — a systematic review. Brain Sciences. 2024;14(1):27. doi:10.3390/brainsci14010027 — [C]
5. Roseberry S, Hirsh-Pasek K, Golinkoff RM. Skype me! Socially contingent interactions help toddlers learn language. Child Development. 2014;85(3):956-970. doi:10.1111/cdev.12166 — [B]
6. Myers LJ, LeWitt RB, Gallo RE, Masnick AM. Baby FaceTime: can toddlers learn from online video chat? Developmental Science. 2017;20(4):e12430. doi:10.1111/desc.12430 — [B]
7. Troseth GL, Strouse GA, Verdine BN, Saylor MM. Let's chat: on-screen social responsiveness is not sufficient to support toddlers' word learning from video. Frontiers in Psychology. 2018;9:2195. doi:10.3389/fpsyg.2018.02195 — [B]
8. Kuhl PK, Tsao F-M, Liu H-M. Foreign-language experience in infancy: effects of short-term exposure and social interaction on phonetic learning. Proceedings of the National Academy of Sciences. 2003;100(15):9096-9101. doi:10.1073/pnas.1532876100 — [B]
9. Hutton JS, Dudley J, Horowitz-Kraus T, DeWitt T, Holland SK. Associations between screen-based media use and brain white matter integrity in preschool-aged children. JAMA Pediatrics. 2020;174(1):e193869. doi:10.1001/jamapediatrics.2019.3869 — [C]
10. Kartushina N, et al. Associations between shared book reading, daily screen time and infants' vocabulary size. Journal of Child Language. 2025. https://www.cambridge.org/core/journals/journal-of-child-language/article/associations-between-shared-book-reading-daily-screen-time-and-infants-vocabulary-size/510AE5663835A8EAE49CF0E51456DA04 — [C]
11. Council on Communications and Media, American Academy of Pediatrics. Media and young minds. Pediatrics. 2016;138(5):e20162591. doi:10.1542/peds.2016-2591 — [guideline: expert consensus]
12. World Health Organization. Guidelines on physical activity, sedentary behaviour and sleep for children under 5 years of age. Geneva: WHO; 2019. — [guideline: expert consensus]
13. Centers for Disease Control and Prevention. CDC's Developmental Milestones: Milestone Moments booklet. https://www.cdc.gov/act-early/media/pdfs/milestone-moments-booklet_eng-print_updated-9.2023_508-p.pdf — [guideline: developmental screening reference]
14. Pempek, Kirkorian & Anderson, Journal of Children and Media 2014 (lab study, n=49: fewer words and new words per minute with background TV on) — [C]
15. Christakis et al., Archives of Pediatrics & Adolescent Medicine 2009 (n=329; -770 adult words/hour of audible TV; four authors employed by LENA, the recording-device company) — [C]
16. McDaniel & Radesky, Child Development 2018 (technoference and fathers: less responsive parenting, more child behaviour problems) — [C]
17. McDaniel & Radesky, Pediatric Research 2018 (mothers: perceived stress/technoference and interaction quality) — [C]
18. McDaniel, Human Behavior and Emerging Technologies 2019 (technoference review) — [C]
19. Radesky et al., Pediatrics 2014 (n=744: difficult temperament predicts later handheld-device use) — [C]
20. Radesky et al., JAMA Pediatrics 2016 (bidirectional TV-language-concern links, 6-14 months) — [C]
21. Sim et al., BMC Pediatrics 2025 (systematic review: parental stress/depression/anxiety track with problematic child screen use) — [C]

Changelog

• 2026-09-07: Topic created (draft).
• 2026-09-07: Second-pass editorial check — added citations to short answer and milestones section; corrected reference 3 journal attribution (Early Child Development and Care, not ECRQ); added CDC milestone reference [13].
• 2026-09-08: Parental-outcomes audit applied: added "What it means for the parents (not just the child)" section plus technoference practical bullet and respite evidence gap; added refs 14–21. Topic rating unchanged (B–C).
• 2026-09-09: Delegated author review (Guille, reviewer of record; agent executed under delegation). Corrected WHO age bands and attributed video-chat exception to AAP only; removed Japanese odds ratios from reader prose (no absolute group risks); replaced "no evidence screens damage the brain directly" with observational-limitations framing and cited Hutton with cross-sectional qualification; softened technoference claims (cross-sectional/bidirectional); moved parental-phone bullet out of professional-help section into practical considerations; replaced "is not a crisis" with "can't estimate the effect of occasional use"; resolved Kirkorian as excluded (same n=49 dataset as Pempek 2014). Status: approved.