Class Sizes
Average primary class size in the OECD is 21 students. In this selection the range runs from Latvia at 14 to Chile at about 31, the largest in the OECD, with Israel, Japan and the UK around 27. Student-teacher ratios (which include all teaching staff, not just classroom teachers) average 15 in primary OECD schools. The research evidence on class-size effects is mixed and modest — small reductions matter less than teacher quality.
Key insights
Class size vs student-teacher ratio differ
Average class size counts students per actual classroom. Student-teacher ratio counts students per all teaching staff (including specialists, support, part-time). Student-teacher ratio is usually about 30% lower than class size (15 vs 21 in OECD primary schools) because not all teachers are in front of students at any one time. Comparisons should specify which is meant — they're not interchangeable.
Larger classes coexist with top scores in East Asia
Japan runs primary classes of about 27, well above the OECD average, and Chinese city schools often run larger, yet both achieve top PISA outcomes. The combination of large classes with high outcomes suggests pedagogical approach, teacher selection, and cultural factors matter more than class size in isolation. Copying East Asian class sizes into Western contexts without other reforms would likely worsen outcomes.
Research evidence is mixed
STAR (Tennessee, 1985-89) — randomized experiment showing meaningful gains from class reductions to 13-17 in K-3. But many other studies show small or null effects from typical class-size variations within the 18-25 range. Meta-analyses suggest meaningful effects below 20 students; little effect at typical OECD ranges. Class-size reduction is expensive — the cost-benefit case is weak unless paired with quality teacher recruitment.
Average primary class size — OECD countries
Students per primary class, OECD Education at a Glance 2024
Show data table
| Avg class size | |
|---|---|
| Chile | 31 |
| Japan | 27 |
| Israel | 27 |
| Mexico | 26 |
| UK | 27 |
| Spain | 21 |
| France | 23 |
| Germany | 21 |
| Italy | 19 |
| OECD avg | 21 |
| USA | 21 |
| Sweden | 20 |
| Finland | 19 |
| Slovenia | 18 |
| Estonia | 17 |
| Latvia | 14 |
Key Finding: Chile has the largest classes, followed by Israel, Japan, the UK and Mexico; Latvia, Estonia and Slovenia have the smallest in this selection.
Primary student-teacher ratio — major economies
Total enrolment / total teaching staff (full-time equivalent)
Show data table
| Pupils per teacher | |
|---|---|
| Tanzania | 44 |
| DRC | 35 |
| Pakistan | 40 |
| India | 32 |
| Brazil | 21 |
| Mexico | 19 |
| World | 23 |
| USA | 14 |
| UK | 14 |
| Germany | 12 |
| Japan | 16 |
| China | 16 |
| OECD avg | 15 |
| Sweden | 12 |
| Norway | 9 |
| Greece | 9 |
Key Finding: OECD averages cluster around 15-16. Emerging markets often have higher ratios — 30+ in parts of sub-Saharan Africa.
Methodology & caveats
Why class size effects might be small
Behavior management gains from smaller classes are real but limited above ~20 students. Individualized attention may not increase linearly with class-size reduction. Teacher quality effects swamp class-size effects in most decompositions. Class-size reduction also raises the marginal teacher (newer, less experienced) which can offset gains.
Practical reform considerations
Class-size reduction is one of the most expensive education reforms — cuts of 5 students per class require 25% more teachers, classroom space, and ongoing salary commitments. Reform programs often produce smaller-than-expected gains because the marginal teacher recruited is less experienced than the average. Teacher quality interventions typically deliver more learning per dollar.
Equity considerations
Where class-size reduction helps most: K-3, especially for disadvantaged students. Where it helps least: secondary, advantaged students. Targeted class-size reduction in early grades for disadvantaged populations has the strongest evidence base. Across-the-board reductions don't focus resources where they matter most.
How class size is measured, and what would change the picture
OECD class size comes from Education at a Glance and counts students per class in regular classes, excluding settings such as special-needs groups where small numbers are the point. A national average also hides wide spread: rural schools in a country with long distances can sit far below the mean while city schools are full, so a parent's experience may differ sharply from the headline. The average has barely moved: the OECD reports primary classes stable at about 21 students between 2015 and 2024. The student-teacher ratio sits beside it but answers a different question, since teachers who do not hold a regular class still count, and countries with more teaching hours per teacher need more classes per teacher. That is why a country can have a low ratio and big classes at once. Two things would shift this picture: falling births, which will shrink classes in much of Europe and East Asia unless schools merge, and evidence from tutoring and AI tools on whether individual attention can be delivered without smaller classes. Tennessee's STAR remains the reference experiment because it randomised pupils and teachers, which most later studies could not.
Cite this page
GlobeChart. “Class Sizes.” Last updated 2026-10-31. https://globechart.com/education/class-size/
Please also credit the original publisher named under each chart: OECD Education at a Glance 2024; World Bank EdStats, OECD Education at a Glance. Spotted an error? Tell us.