What Age Do Kids Actually Start School Around the World?

Someone in your group chat just sent the Finland thing again. Kids there don't start school until 7, it says. American 5-year-olds are already at desks. Here's the annoying part: that's true. Finnish kids really do start primary school the year they turn 7. So do Swedish kids. So do Estonian kids.

On this page▾

The three clocks, country by countryWhat does that compulsory year actually look like, room by room?California just built its own version of this splitDoes starting later actually help kids? The research disappoints both sidesChina just legislated the pressure back outSo what does this mean for choosing a school here

It's still almost never the whole story. "Starting school" is really at least three separate events, bundled into one word: the age a child is legally required to be somewhere, the age primary instruction actually begins, and the age most kids first show up to some kind of group care. Formal academic instruction — actual reading lessons — can start at yet another point again, depending on the curriculum. Every one of these is real. Mash them into a single "starts at 7" line and you get a fact with the wrong shape.

Take Finland seriously instead of turning it into a vibe. Finnish law has required every 6-year-old to participate in pre-primary education — or a corresponding alternative activity that meets the same objectives — since 2015. That's four to five hours a day, with complementary ECEC available for families needing a full day. Basic education — the thing that's actually "primary school" — starts the following year, at 7. Plenty of Finnish kids are in some form of early childhood care well before that six-year mandate kicks in. So the real question is what kind of room a Finnish 4- or 5-year-old spends the day in, and what the adults there ask of them.

The three clocks, country by country

Ontario and California aren't countries. The same split shows up in both, so they're in the table too. "Compulsory from" is the earliest age a law requires something of a family — attendance in most places, instruction alone in a few (France allows home instruction instead of physical attendance — since 2022, only with state authorization on a narrow set of grounds, not by simple declaration). "Primary (ISCED 1)" is when real primary school — what most countries classify as ISCED 1 — actually starts. "Before then" is what's typical before either of those dates, whether or not a law requires it.

Compulsory-something vs. real primary school, eight systems. Full details for all fifteen in the table below.

Read that table by column, not by row, and the meme's real failure shows up. Compulsory education and primary-school entry sit at the same age in most of these places. That coinciding age is almost never 7. Finland and Sweden pair primary school at 7 with compulsory pre-primary at 6. Estonia also begins compulsory primary education at 7, but preschool before then remains optional. None of these systems assumes that primary school at 7 means children receive no organized education before then. They're deciding, differently, which of the three clocks counts as "school."

What does that compulsory year actually look like, room by room?

The table shows dates. It doesn't show what a kid actually does between 9 and 3, or the paper trail behind why a ministry somewhere picked that particular age. Here's the material version of the eight rows in the chart above: what the room looks like at each system's earliest required or typical age, and what's actually on the record about why it landed there.

France — required at 3, school at 6

France's law reaches furthest down, to age 3, and there's an unusually clear paper trail for why. École maternelle runs a 24-hour instructional week — the schedule set by Article D. 521-10 of the Code de l'éducation, in force since 2013 — and the national curriculum leads with one thing: language — talking, naming, listening, retelling a story — alongside play, art, and physical activity. The education ministry's own account of the 2019 law names what maternelle is actually for: it "develops the skills indispensable" for a child to later tackle, at école élémentaire and age 6, "the fundamentals — reading, writing, counting."

Near-universal attendance came decades before the law did. France has run something close to universal maternelle since 1990, per a demographic review of INSEE enrollment data; by 2018, the year before the law changed anything, 96.6% of French 3-year-olds were already enrolled. The 2019 loi pour une école de la confiance had a different job: the ministry's own published account frames it as a targeted equity move, aimed at reducing inequality "from the youngest age," and it names language specifically as "the first of these inequalities" the law was meant to close. The honest coda comes from the government's own later evaluation: the law's effect on enrollment was "limited, if not nil" — attendance barely moved, because it was already this high. What "compulsory" added, on top of an attendance rate that barely needed the push, was a legal duty: enrollment, instruction and attendance became enforceable, with oversight behind them, for the small number of families who weren't already there.

Netherlands — a building merger, not a law

The Netherlands built the same idea into a building instead of a law. Until 1985, kleuterschool (nursery school) and lagere school (primary school) were separate institutions. The Wet op het Basisonderwijs merged them into one basisschool spanning groups 1 through 8, ages 4 to 12. Statistics Netherlands' own account of the change names the aim directly: easing the jump from kleuterschool into lagere school, replaced with "an unbroken learning path" for the whole span. Groep 1–2, the merged system's youngest rooms, still count as pre-primary today — most Dutch kids are in the building at 4, a year before compulsory education starts at 5 and two years before real primary school, groep 3, begins at 6.

Denmark — a play year made compulsory in 2009

Denmark's børnehaveklasse — Grade 0 — runs 27.8 hours a week, about 1,110 hours a year. The education ministry describes the method plainly: children experience school as a stable setting for development "through play-based activities," a bridge between daycare's play and Year 1's subject-based lessons. It became compulsory in 2009, when a broader school reform stretched Denmark's compulsory-education span from nine years to ten. Before that, municipalities had been required to establish a børnehaveklasse since 1980 — attendance just wasn't required.

Finland — voluntary until 2015

Finland's compulsory pre-primary year — the four-to-five-hour day already described above — was voluntary until a 2015 change to the Basic Education Act made it compulsory nationwide, aimed at reaching the small share of six-year-olds a voluntary system was still missing. Pre-primary runs on the same national curriculum whether it happens inside a school building or an early-childhood center. The setting varies. The year, and what it covers, doesn't.

Sweden — the fullest paper trail

Sweden's case is on the record in more detail than any other system here. The government's own bill, Skolstart vid sex års ålder ("School start at age six"), argued that "all children have the right to a good and equal foundation" before primary school, and that roughly 2% of six-year-olds — about 2,000 kids a year — were falling through the cracks of a voluntary system, showing up in none of förskoleklass (the preschool class), förskola (preschool), or first grade. The bill names a second reason too: making the year compulsory lets teachers "discover early the children who may need extensive additional support." It took effect that autumn, 2018.

New Zealand — no fixed date at all

New Zealand runs on a different clock. Kids have long started school "continuously," any school day on or after their fifth birthday, rather than on one shared date. A 2017 law gave schools the option to switch to cohort entry instead: fixed start dates, up to eight a year, so new entrants arrive as a group. It's still a minority choice — 7% of schools had adopted it, per the Education Review Office's own count. Among the schools that did switch, almost three-quarters of principals and new-entrant teachers said cohort entry had social and emotional wellbeing benefits — and more specifically, most said the children themselves settled into class better: 93% of principals, 79% of teachers.

Estonia — optional, backstopped by an assessment

Estonia keeps the oldest primary-start age on this list — the "obligation to learn" begins at 7. Municipalities must guarantee a preschool-institution place for any child from 18 months to 7 whose family wants one; parents of under-4s can choose a separate childcare arrangement instead. The preschool institutions run on a national curriculum, most recently set by a 2008 government regulation. None of it is compulsory — attendance is on request. A child who finishes the preschool curriculum is issued a school-readiness card for whatever primary school comes next; a child who didn't attend gets a readiness assessment from a regional advisory centre instead. Finland and Sweden require the room. Estonia leaves it optional and backstops the choice with an assessment.

California — a fix that outgrew its cutoff

California's own split has a paper trail too. Transitional kindergarten began narrowly, as a fix. The 2010 Kindergarten Readiness Act (SB 1381) moved the kindergarten birthday cutoff from December 2 to September 1, phased in one month a year for three years. TK was built into that same statute for exactly the kids the shifting cutoff displaced — a child "who will have his or her fifth birthday between September 2 and December 2," who would have qualified for kindergarten under the old date, gets a seat in TK instead, in "the first year of a two-year kindergarten program." That narrow cohort didn't hold for long. A 2015 law gave districts the local option to admit even younger four-year-olds into TK if they chose to — well short of a state mandate, but the first crack in the original cutoff. It took a 2021-22 state budget to turn TK into something else entirely: a universal entitlement for every four-year-old, finished by 2025-26.

California just built its own version of this split

This isn't only a story about Northern Europe. Since the 2025–26 school year, every California 4-year-old (turning 4 by September 1) is entitled to free transitional kindergarten. It's optional. Compulsory school age is still 6, unchanged, under Education Code § 48200. So California now has its own three-way split. There's an age when the state entitles every child to a free public seat (4). An age most kids actually show up (5, for kindergarten). And an age the law requires it (6). That's the same shape as Finland's 6/7 gap, or the Netherlands' 4/5 gap: a state deciding that "available" and "required" don't have to be the same number. Our pillar guide to Pre-K, TK and preschool in California covers what these programs actually cost and who qualifies. Our piece on California's compulsory-age law covers the exceptions to the age-6 rule itself, including the one most guides miss.

Does starting later actually help kids? The research disappoints both sides

This is where the meme picks up its emotional charge. There's the idea that a later start produces calmer, smarter kids. There's the opposite idea, that starting early gives a lasting edge. Four different studies bear on this, from four different countries. None of them hands either side a clean win.

The strongest evidence for "later helps" is Danish. Economists Thomas Dee and Hans Henrik Sievertsen used the exact age cutoff for Danish kindergarten entry. They compared kids born just before it against kids born just after. A one-year delay cut inattention and hyperactivity, as rated by mothers on a standard behavior scale, by a large margin at age 7. That finding comes with real limits worth knowing before you lean on it. The result was identified mainly for girls. Boys didn't comply with the entry cutoff consistently enough to measure. The families in the sample were also wealthier, more educated, and less likely to be recent immigrants than Denmark's population as a whole. And the "extra year" wasn't academic prep, or a year at home. It was a year in Danish public childcare, which leans heavily play-based. The authors themselves say their design can't isolate an effect on grades or test scores at all — age and time in school move together once you fix the birth year. This is a real result about Danish girls in Danish play-based care. It's not a study of what happens if an American family holds a kid back a year.

Pull back to a bigger, population-scale study and the picture tilts the other way. Economists Sandra Black, Paul Devereux and Kjell Salvanes used large national administrative cohorts in Norway. Their IQ measure came from scores at military conscription, around age 18 — outside school, and men only, since that's who Norway conscripts — which let them separate age-at-test from starting age in a way in-school testing can't. They found a small positive effect of starting school younger on those men's IQ scores. Across both boys and girls, starting younger produced almost no effect on how much schooling people finished, and an early-earnings edge that faded to nothing by age 30. It isn't a clean win for starting early either: for girls specifically, starting younger also raised the odds of teenage pregnancy in the same data. Whatever this study proves, it isn't "wait as long as possible." It also isn't "earlier is obviously better."

Then there's a claim you'll sometimes see, that countries with later school-starting ages post better PISA reading scores. It has a real source, and the source doesn't support it. A 2009 analysis by Sebastian Suggate re-ran the numbers across 55 countries in PISA 2006. It found no significant link between a country's school-entry age and how its 15-year-olds read, once wealth and social factors were controlled for. A country's starting age, on its own, doesn't predict how well its teenagers read.

The last piece is about reading instruction specifically — it's the study Waldorf schools' "wait on reading" approach often gets defended with. Suggate and colleagues compared New Zealand children who started formal reading instruction at 5 against a Rudolf Steiner (Waldorf) cohort that started about 19 months later, closer to 6 and a half than a clean 7. The early group read better at first. No surprise there. By age 11, that gap had closed. A separate, smaller comparison of middle-school-age children found the later group with a slight edge in reading comprehension, though not in fluency. The authors are careful about their own finding. It's worth quoting them directly: "we do not interpret the current findings as evidence that no reading instruction should occur before age seven." And on the design itself, a non-experimental comparison of two selected, relatively advantaged samples: it "cannot generalize to other, less fortunate, contexts." An older 2002 UK review of earlier international data landed in a similarly unresolved place: "a lack of conclusive evidence concerning the benefits of starting school at different ages."

Put the four together and nobody gets to be smug. The Danish study found a large behavior effect, mainly for girls, but could not identify the effect on academic achievement. Starting younger produced little lasting difference in how much schooling people finished or what they earned by 30, for boys and girls alike. Among women, it also raised the odds of teenage childbearing. A country's calendar doesn't predict its teenagers' reading scores. The selected New Zealand samples showed no lasting reading disadvantage by age 11, but the design cannot tell us whether delaying instruction caused that result or whether it would generalize.

China just legislated the pressure back out

If your mental picture of China is the most academically pushed early-years system anywhere, that's been fair. It might still be true in practice. But in November 2024, China's legislature passed a national Preschool Education Law. It took effect June 1, 2025, and it does something the US has no national equivalent of: it bans kindergartens from teaching the primary-school curriculum. Activities have to match "the developmental stages and characteristics of preschool children" instead. A "double reduction" policy in 2021 already cut homework and restricted outside tutoring for young children. From autumn 2025, public kindergartens started waiving fees for the year right before primary school. Private kindergartens must cut fees by a matching amount. The country most associated with early academic pressure has spent four years legislating some of it back out — on the curriculum side, and on the cost side.

So what does this mean for choosing a school here

Not much directly. This is an international curiosity, not a California decision. But it's the same shape as the exceeded-vs-met split we write about constantly: one number on a label tells you less than what's actually happening in the room. "Starts at 7" and "starts at 3" aren't opposite philosophies with a winner. They're different countries drawing the line between daycare and school in different places, usually while running some kind of group care well before either line.

If you're deciding when your own kid starts something in California — TK, kindergarten, or holding them back a year — California's legal floor of 6 matches Finland and Sweden's compulsory pre-primary year, not their primary start, and the primary-school starting age in Germany, Japan, China and Ontario. Kindergarten itself isn't even required by law here. If the redshirting question is what actually brought you to this page, the honest answer sits above. The evidence is genuinely mixed, split by country and by outcome. Nobody's data hands you a clean rule for your specific kid.