What PISA 2025 Really Tells Us About Estonia’s Education Model

High performance, equity and digital maturity — but also a warning about reading, cognitive effort and the age of AI

By Mehmet ALTAYFounder & International Projects Director, Teachers Education Academy

Estonia enters the PISA 2025 debate from a position that very few education systems can claim. It is not simply a high-performing country in one subject, nor is its international reputation based only on digitalisation. According to the OECD’s PISA 2025 Results (Volume I): Future-Ready Students, Estonia remains among the top-performing education systems in science, reading and mathematics simultaneously.

Yet the new results deserve a more careful reading than another headline celebrating Estonia’s position in international rankings. PISA 2025 shows both the continuing strength of the Estonian model and the pressures now emerging within it. The most important question is therefore no longer simply why Estonia performs well. It is what Estonia must protect, strengthen and rethink as education enters a period increasingly shaped by artificial intelligence, digital distraction and changing patterns of student learning.

Estonia remains a high-performing system

Estonian students achieved 527 points in science, 499 in reading, 508 in mathematics and 541 in computational problem solving in PISA 2025. Between 2022 and 2025, science performance remained essentially stable, increasing by one point, mathematics declined by two points, while reading decreased by 12 points.

These figures matter because Estonia is not succeeding narrowly. Its students continue to perform strongly across multiple domains, and the country remains one of the few systems capable of combining traditional academic achievement with strong performance in computational problem solving.

The science result is particularly significant. Estonia scored 527 points and remains among the world’s strongest systems in this domain. Even more importantly, the OECD identifies Estonia as one of the above-average-performing systems with comparatively limited variation in student science performance.

This suggests that Estonia’s strength is not based solely on producing a small group of exceptionally high-performing students. The system is also relatively successful in limiting very low performance.

Indeed, fewer than 15% of Estonian students perform below the baseline Level 2 in science. Estonia belongs to a very small group of systems achieving this result, alongside B-S-J-Z (China), Macao, Singapore, Japan, Chinese Taipei and Korea.

That is one of the most important features of the Estonian model: high performance is combined with relatively broad participation in that performance.

Estonia’s real science strength is deeper than subject knowledge

PISA 2025 also provides a more detailed picture of what Estonian students can actually do in science.

Estonia scored 521 in explaining phenomena scientifically, 530 in constructing and evaluating designs for scientific enquiry, and 528 in researching, evaluating and using scientific information.

The OECD specifically identifies Estonia, together with B-S-J-Z (China) and Macao, as having a relative strength in constructing and evaluating designs for scientific enquiry.

This is an important distinction.

A successful science education system should not only enable students to remember scientific facts. Students need to ask questions, investigate, evaluate evidence, judge the quality of information and understand how knowledge is produced.

Estonia’s PISA profile suggests strength precisely in these areas.

For educators, this reinforces the value of inquiry-based learning, experimentation, problem solving, interdisciplinary STEAM approaches and pedagogies that require students to actively construct knowledge rather than passively receive it.

Computational problem solving may be one of Estonia’s most important advantages

Estonia’s performance in computational problem solving deserves particular attention.

With 541 points, Estonia is far above the OECD average of 500 and belongs to the strongest-performing OECD countries in this domain. The report places Australia, Estonia, Japan, Korea and New Zealand among OECD countries scoring above 530.

Estonia’s result is close to Hong Kong (China), which scored 545, and above Korea at 538.

This finding helps correct a common misunderstanding about digital education.

Digital maturity should not be measured simply by the number of devices in classrooms, the number of digital platforms used, or the frequency with which students interact with technology.

What matters is whether students can think, reason, solve problems and make decisions in technology-rich environments.

Estonia’s result suggests that its digital education strength is fundamentally cognitive, not merely technological.

That distinction will become increasingly important as artificial intelligence becomes more deeply integrated into education.

Estonia’s greatest warning sign is reading

The most concerning part of Estonia’s PISA 2025 profile is reading.

Between 2022 and 2025, Estonia’s reading score declined by approximately 12 points, while science remained stable and mathematics changed only slightly.

The longer-term picture deserves even more attention. Over the decade covered in the report, Estonia recorded declines in science, reading and mathematics, with reading showing a particularly notable deterioration.

This cannot simply be dismissed because Estonia remains internationally competitive.

A high-performing system can still experience erosion.

The OECD itself frames declining reading performance as a major challenge for contemporary education. It highlights the importance of students being able to evaluate information, connect information from different sources and think critically about what they read.

These abilities are becoming more—not less—important in the age of generative AI.

Students are now surrounded by instantly generated summaries, explanations, answers and recommendations. In such an environment, reading is no longer only about understanding written text. It is increasingly connected to information judgement, source evaluation, critical thinking and intellectual independence.

For Estonia, protecting strong digital competence while rebuilding deep reading may therefore become one of the defining educational challenges of the coming decade.

Estonia demonstrates that excellence and equity can coexist

Perhaps the most impressive Estonian finding in PISA 2025 concerns equity.

OECD analysis places Estonia among a small group of education systems that combine high inclusion, socio-economic fairness and strong academic performance.

Estonia is among the systems where more than 75% of students reach at least Level 2 in science, less than 10% of variation in science performance is associated with socio-economic status, and overall performance remains above the OECD average.

The OECD identifies Estonia alongside Canada, Ireland, Japan, Korea, Macao and B-S-J-Z (China) as systems combining socio-economic fairness with strong performance.

This matters enormously.

The purpose of a strong education system cannot simply be to maximise average scores. A system should also minimise the extent to which family background determines educational opportunity.

Estonia continues to demonstrate that these goals are not mutually exclusive.

High achievement does not require accepting large inequalities between students.

Educational resilience remains an Estonian strength

The PISA 2025 analysis also places Estonia among the systems with notable numbers of academically resilient students—students who are socio-economically disadvantaged but nevertheless achieve strong academic outcomes relative to peers in their own country.

This adds an important dimension to the equity story.

A resilient education system does not merely perform well under favourable conditions. It gives students a realistic chance of succeeding even when their starting circumstances are less advantageous.

For European education policy, this may be one of the most transferable lessons from Estonia.

The most interesting question for other countries should not be, “How can we copy Estonia’s technology?”

It should be:

How does Estonia organise teaching, schools, support and expectations so that socio-economic disadvantage has a comparatively limited effect on student outcomes?

Estonia and artificial intelligence: a surprising finding

One of the most interesting findings in the report concerns AI use.

Despite Estonia’s international reputation as a highly digital society, the OECD reports that students in Estonia use AI for schoolwork less frequently than students in several other high-performing systems.

The report contrasts Korea and Singapore, where digital and AI policies are comparatively comprehensive, with Japan and Estonia, where AI is used less frequently for schoolwork and digital-learning policies are described as comparatively less comprehensive in scope.

This finding should challenge the assumption that educational innovation means maximising AI use.

Estonia’s results suggest that a digitally advanced education system does not need to become an AI-intensive education system overnight.

That distinction is important.

The objective should not be:

more AI.

The objective should be:

better learning.

AI should be introduced where it strengthens learning, feedback, creativity, problem solving and access to support—not simply because the technology exists.

AI should support thinking, not replace it

PISA 2025 delivers a particularly relevant message for schools developing AI strategies.

Across OECD countries, moderate educational use of digital devices is associated with stronger outcomes, while excessive use and digital distraction are associated with weaker performance.

The OECD makes an equally important argument about artificial intelligence: technology should enhance the intellectual work involved in learning rather than bypass it.

The report summarises the principle powerfully: AI should become “a scaffold, not a crutch” and “a tool for thinking, not a substitute for it.” 

For Estonia, this may be one of the most important messages of PISA 2025.

The next phase of digital education should not be defined primarily by technological adoption.

It should be defined by pedagogical judgement.

Teachers and schools will increasingly need to decide when AI adds value, when students should work independently, when technology helps develop understanding and when it removes the cognitive struggle through which genuine learning occurs.

Family support is another positive Estonian signal

While family support declined between 2022 and 2025 in most participating systems, Estonia moved in the opposite direction.

The OECD reports improvements in perceived family support in Estonia, Poland, Viet Nam and Iceland.

This is another reminder that educational performance should not be interpreted only through curriculum, technology or assessment.

Learning is shaped by an ecosystem.

Teachers matter. Schools matter. Students’ motivation matters. But families and the quality of relationships surrounding learners matter as well.

Estonia’s future education strategy therefore needs to continue considering learning as something created through the interaction of school, teacher, student and family, rather than through classroom interventions alone.

Estonia should celebrate PISA 2025 — but not become complacent

PISA 2025 gives Estonia many reasons for confidence.

It remains internationally strong in science, mathematics and reading. It performs exceptionally well in computational problem solving. It combines high performance with relatively strong socio-economic fairness. Its students demonstrate particularly strong capabilities in scientific enquiry.

But the report also contains clear warnings.

Reading performance is declining. Longer-term mathematics trends require attention, especially because the OECD associates Estonia’s decade-long mathematics decline with deterioration among lower-achieving students, contributing to a widening performance gap.

At the same time, schools are entering an era in which AI can either strengthen learning or gradually replace some of the intellectual processes through which students learn to think.

That is why Estonia’s next educational challenge is probably not digitalisation itself.

Estonia has already demonstrated that it can build a digitally capable education system.

The new challenge is more demanding:

How can Estonia preserve deep learning, reading, critical thinking, intellectual effort and human agency while taking advantage of artificial intelligence?

PISA 2025 suggests that the future of successful digital education will not belong to the systems that use the most technology.

It will belong to the systems that understand when, why and how technology improves human learning.

Estonia is exceptionally well positioned to lead that next conversation.

Mehmet ALTAYFounder & International Projects DirectorTeachers Education Academy

This analysis was prepared by Teachers Education Academy based on OECD (2026), PISA 2025 Results (Volume I): Future-Ready Students. Interpretations and conclusions presented in this article are those of Teachers Education Academy and should not be understood as official OECD positions.

#PISA2025 #EstoniaEducation #FutureOfEducation #EdReform #DigitalEducation #EducationModel #TeacherAutonomy #EdTech