PISA Scores Comparison: Math, Reading, Science by Country

PISA 2022 is the latest full international benchmark for how 15-year-olds use mathematics, reading, and science in applied settings rather than in narrow recall tests. The cycle covered about 690,000 students representing roughly 29 million learners across 81 countries and economies, and mathematics was the main focus area in the delayed 2022 round. For any country comparison, the visible ranking is only the first layer. Under it sit proficiency thresholds, subject balance, long-run movement, and equity gaps. [Source-1✅] [Source-2✅]

How to Read a Country Comparison

PISA reports the three core domains separately. That matters. A country can look stronger in one area and merely solid in another. The United States, for example, sits well above the OECD mean in reading but below it in mathematics, while several East Asian systems occupy the top end in all three subjects. A plain rank list hides that subject profile. [Source-5✅]

There is also a dataset-count issue that many comparison pages skip. The OECD’s 2022 documentation refers to 81 participating countries and economies, while the subject ranking tables cited below display 78 published country/economy entries. When different websites show different counts, this is usually the first reason. It does not change the OECD averages, but it does affect how a public ranking page looks. [Source-6✅] [Source-7✅] [Source-8✅]

• Mean score shows the average position of a system.
• Level 2 proficiency shows how many students reach a baseline that OECD treats as the minimum needed for further study and daily problem-solving.
• Level 5 and 6 shows how deep the top end of performance really is.

2022 Rankings in Mathematics

The table below uses the published 2022 mathematics scores shown in the cited PISA/OECD-derived ranking page. [Source-9✅]

Mathematics is where the concentration at the top looks sharpest. Six of the top six positions belong to East Asian systems, then come Estonia and Switzerland, followed by Canada and the Netherlands. The OECD also notes that only 16 of the 81 participating systems had more than 10% of students at Level 5 or 6 in mathematics, while across OECD countries 69% reached at least Level 2. That means a country can have a decent average and still have too thin a high-performance layer or too many students below baseline. [Source-10✅]

Another detail matters. OECD country-note evidence shows that more than 85% of students in Singapore, Macao, Japan, Hong Kong, Taiwan, and Estonia reached at least Level 2 in mathematics. So the story is not only about very high averages. It is also about broad baseline mastery across the student body. [Source-11✅]

2022 Rankings in Reading

The reading ranking below uses the published 2022 scores shown in the cited PISA/OECD-derived page. [Source-12✅]

Reading is more spread across regions than mathematics. Singapore remains first, but Ireland, Japan, Korea, Taiwan, Estonia, Canada, the United States, and New Zealand all cluster within a much tighter band than the mathematics leaders do. That is why a tiny change in rank near the top can be less informative than the underlying proficiency distribution. OECD reports that 74% of students across OECD countries reached Level 2 or higher in reading; country notes also show that the share ranged from 89% in Singapore to 8% in Cambodia. Wide spread, even when the upper ranks look crowded. [Source-13✅]

Ireland’s position is especially notable because it sits near the summit in reading while still posting mathematics and science scores above the OECD means. The United States shows a different profile: strong reading, solid science, but mathematics below the OECD benchmark. Read the three domains together, and the picture becomes much cleaner. [Source-14✅] [Source-15✅]

2022 Rankings in Science

The science ranking below uses the published 2022 scores shown in the cited PISA/OECD-derived page. [Source-16✅]

Science again places Singapore at the top, with Japan, Macao, Taiwan, Korea, and Estonia close behind. Then come Hong Kong, Canada, Finland, and Australia. The OECD average in science was 485, and 76% of students across OECD countries reached at least Level 2. Singapore’s country note adds another useful layer: 24% of its students were top performers in science, against an OECD average of 7%. That is not just a lead. It is a very deep upper tail. [Source-17✅]

What the Three Tables Show When Read Together

• Singapore is the only system that clearly leads all three domains at once.
• Japan and South Korea pair very high mathematics scores with equally strong reading and science.
• Estonia is the standout European performer, staying inside the top group across all three subjects.
• Canada is unusually balanced: top 10 in reading and science, and top 10 just outside the very highest mathematics cluster.
• Ireland is a reading leader with mathematics and science still above the OECD means.

This cross-subject view is where many simple ranking pages stop too early. A country may shine in one domain but not deliver the same strength across the board. Another may never reach the very top in a single subject, yet remain consistently strong in all three. For school-system comparison, that balance is often more revealing than a single headline rank. [Source-18✅]

What the Raw Rankings Miss

Why Proficiency Bands Matter More Than Tiny Rank Gaps

A one-point or two-point shift in rank can look dramatic on a social graphic. In practice, the more useful question is how many students clear baseline proficiency and how many reach the advanced band. OECD’s own reporting keeps returning to Level 2 and Level 5/6 for exactly that reason. In mathematics, more than 10% of students reached Level 5 or 6 in only 16 of 81 systems, while less than 5% did so in 42 systems. The spacing between rank positions is not the whole story. [Source-19✅]

Equity Still Separates High-Score Systems

The OECD reports that socio-economically disadvantaged students were seven times more likely, on average, than advantaged students not to reach basic mathematics proficiency. Even so, 10% of disadvantaged students across OECD countries still placed in the top quarter of mathematics performance within their own countries. That combination of achievement and mobility is one of the most useful markers in the whole dataset, and it often disappears when country comparison is reduced to a neat vertical ranking. [Source-20✅]

OECD materials also point to a smaller set of systems that pair strong proficiency in mathematics, reading, and science with stronger socio-economic fairness. That detail changes the interpretation of high scores. A country with a high mean and narrower social stratification is telling a different story from a country with the same mean but a much steeper social gradient. [Source-21✅]

The 2018-2022 Shift Changed the Map

The latest cycle was not business as usual. OECD average performance fell by a record 15 points in mathematics and by 10 points in reading between 2018 and 2022, while science did not show the same average drop. That means a current country comparison has to be read against a changed baseline, not against the calmer pattern that many readers still associate with the 2012 or 2018 cycle. [Source-22✅]

Not every system moved the same way. OECD highlights Japan, Korea, Lithuania, and Taiwan as systems that managed to maintain or improve learning outcomes, fairness in the distribution of learning opportunities, and student well-being during this difficult period. It also notes that Colombia, Macao, Peru, and Qatar improved in all three subjects on average since they first joined PISA. Add Türkiye to the long-run picture and you get another reminder: the map is not static. Some countries are still climbing, even when the global mood focuses on decline. [Source-23✅]

What a Serious Country Comparison Should Keep in View

A useful PISA scores comparison does not stop at “who is first.” It reads mathematics, reading, and science as three linked but distinct measures; it watches Level 2 and Level 5/6, not only mean score; it checks whether a system is balanced across subjects; and it looks at whether the score distribution is socially broad or narrowly concentrated. Read that way, the 2022 results show a clear pattern: a small group of systems combine very high averages with deep proficiency and steadier equity, while many others show mixed subject profiles or are still moving upward from lower starting positions. That is the point where a ranking becomes informative rather than decorative. [Source-24✅] [Source-25✅]