Programme for International Student Assessment 2025: national report for Scotland

5 Performance in mathematics

5.1 How mathematics is assessed

The mathematics assessment in PISA 2025 uses the mathematics framework introduced in PISA 2022, when mathematics was the major assessment domain. The same framework was retained for PISA 2025, with mathematics assessed as a minor domain.

The framework is based on the concept of mathematical literacy: an individual’s capacity to reason mathematically and to formulate, employ and interpret mathematics to solve problems in a range of real-world contexts. Rather than assessing mastery of school curricula, PISA focuses on how well 15 year olds can apply mathematical knowledge and skills to situations they may encounter in everyday life.

The framework is organised around three mathematical processes - formulating, employing, and interpreting and evaluating mathematics - which reflect the stages involved in solving real-world problems. Mathematical knowledge is assessed across four content categories: change and relationships, space and shape, quantity, and uncertainty and data. Problems are presented in four contexts -personal, occupational, societal and scientific - to reflect the diverse situations in which mathematics is used.

The framework also places greater emphasis on mathematical reasoning, recognising the increasing importance of critical thinking, problem solving and computational thinking in a technology-rich world.

5.2 Scotland’s performance in mathematics

Scotland’s average score in mathematics in PISA 2025 was 467 points. This was not significantly different from Scotland’s score of 471 points in PISA 2022.

Over the longer-term, Scotland’s mathematics performance has declined. Scotland achieved an average score of 524 points in 2003, when mathematics was first assessed as the major domain. Performance in 2025 was significantly lower than in every assessment cycle between 2003 and 2018.

Scotland’s mathematics performance in 2025 was not significantly different from 2022. Overall, the results therefore show a long-term decline in mathematics performance, but no evidence of a statistically significant change between 2022 and 2025.

5.3 Mathematics performance by sex

In PISA 2025, boys in Scotland achieved a mean mathematics score of 475 points, compared with 459 points for girls. Boys’ mathematics performance was significantly higher than girls’.

Among boys, the mean mathematics score was 475 points in 2025. This was significantly lower than boys’ performance in assessment cycles between 2006 and 2018, but was not significantly different from their performance in 2022.

Among girls, the mean mathematics score was 459 points in 2025. This was significantly lower than girls’ performance in previous assessment cycles between 2006 and 2018, but was not significantly different from their performance in 2022.

Overall, mathematics performance in 2025 was significantly lower than in earlier PISA cycles for both boys and girls. However, for both boys and girls performance in 2025 was not significantly different from 2022.

5.4 Mathematics performance across the United Kingdom

As in previous PISA cycles, there was variation in mathematics performance across the four UK nations.

Scotland achieved a mean mathematics score of 467 points, compared with 488 points for the United Kingdom as a whole. Scotland’s performance was significantly lower than the UK average and England (492 points), significantly higher than Wales (456 points), and not significantly different from Northern Ireland (473 points).

The difference between Scotland and England was evident for both boys and girls. Among girls, Scotland’s mean mathematics score was 459 points and was significantly lower than the mean score for girls in England (478 points). Scotland’s performance among girls was not significantly different from Northern Ireland (469 points) or Wales (450 points).

Among boys, Scotland’s mean mathematics score was 475 points and was significantly lower than the mean score for boys in England (504 points). Scotland’s performance among boys was not significantly different from Northern Ireland (478 points) and was significantly higher than Wales (462 points).

Overall, Scotland’s mathematics performance was significantly lower than England and the UK average, significantly higher than Wales, and not significantly different from Northern Ireland. The pattern by sex was somewhat different: both boys and girls in Scotland performed significantly below their counterparts in England and at a similar level to those in Northern Ireland, but only boys in Scotland performed significantly above their counterparts in Wales.

5.5 How Scotland’s mathematics performance compares internationally

Across all pupils, Scotland achieved a mean mathematics score of 467 points. This was statistically similar to the OECD average of 463 points.

Scotland’s mathematics performance was significantly lower than 14 countries and UK administrations, including Australia, Poland, and Ireland. Japan, Korea and Estonia were the highest-performing countries.

Scotland performed at a statistically similar level to 15 countries and UK administrations, including Finland, Germany, and Sweden. Scotland’s performance was also not significantly different from the OECD average.

Scotland performed significantly higher than 11 countries and UK administrations, including Iceland, Spain, and Israel.

Table 5.1: OECD countries and UK administrations, higher than, similar to and lower than Scotland in mathematics, PISA 2025

Higher score than Scotland

• Australia (478)
• Austria (477)
• Belgium (480)
• Canada (485)
• Czechia (477)
• England (492)
• Estonia (508)
• Ireland (480)
• Japan (525)
• Korea (522)
• Netherlands (483)
• New Zealand (480)
• Poland (484)
• Switzerland (499)
• United Kingdom (488)

Similar score to Scotland

• Denmark (471)
• Finland (469)
• France (458)
• Germany (464)
• Hungary (459)
• Italy (468)
• Latvia (460)
• Lithuania (470)
• Northern Ireland (473)
• OECD average (463)
• Portugal (460)
• Scotland (467)
• Slovak Republic (469)
• Slovenia (460)
• Sweden (464)
• Türkiye (462)
• United States (463)

Lower score than Scotland

• Chile (403)
• Colombia (381)
• Costa Rica (387)
• Greece (424)
• Iceland (450)
• Israel (433)
• Luxembourg (458)
• Mexico (388)
• Norway (452)
• Spain (457)
• Wales (456)

Mathematics performance also declined significantly across the OECD between 2022 and 2025, with significant falls recorded in around half of participating systems with comparable data. Over the past decade, the OECD average has fallen substantially, with declines particularly pronounced among lower-performing pupils and a widening gap across the performance distribution. While Scotland did not record a significant change between 2022 and 2025, its longer-term trajectory and relatively large proportion of pupils below Level 2 are consistent with the wider concern about weakening mathematics performance, particularly at the lower end of the distribution.

5.6 International comparisons by sex

The pattern of international comparisons differed somewhat between boys and girls.

Among girls, Scotland achieved a mean mathematics score of 459 points, which was statistically similar to the OECD average for girls (456 points). Scotland’s performance among girls was significantly lower than 13 countries and UK administrations, including Australia, Ireland, and Belgium; statistically similar to 19, including Austria, Denmark, and Finland; and significantly higher than 8, including Iceland, Greece and Luxembourg.

Among boys, Scotland achieved a mean mathematics score of 475 points, which was statistically similar to the OECD average for boys (470 points). Scotland’s performance among boys was significantly lower than 12 countries and UK administrations, including Australia, Ireland, and Austria; statistically similar to 15, including Finland, Czechia, and Denmark; and significantly higher than 13, including Iceland, Spain and Wales.

Scottish boys occupied a slightly stronger relative international position than Scottish girls. Fewer countries outperformed Scotland among boys than among girls, and Scotland outperformed a larger number of countries among boys. However, for both sexes, performance remained broadly in line with the OECD average and below that observed in the highest performing education systems.

5.7 Mathematics performance by proficiency level

PISA reports performance using a series of proficiency levels that describe the types of mathematical tasks pupils can typically complete. Pupils performing at Level 2 or above are considered to have achieved the baseline level of mathematical proficiency required to participate effectively in society. Pupils performing below Level 2 are more likely to face difficulties applying mathematical knowledge and skills in everyday situations. Pupils performing at Level 5 or above are considered high performers.

In 2025, 32.0% of Scottish pupils performed below Level 2. This was significantly higher than in every assessment cycle between 2006 and 2018, but was not significantly different from 2022 (30.7%). This means that almost one third of pupils did not reach the baseline level of mathematical proficiency in 2025.

At the upper end of the proficiency scale, 7.2% of pupils performed at Level 5 or above in 2025. This was significantly lower than in 2006, 2009, 2012 and 2018, but was not significantly different from 2015 or 2022.

5.7.1 International comparison of mathematics by proficiency levels

In 2025, 32.0% of pupils in Scotland performed below Level 2 in mathematics. This was not significantly different from the OECD average, Northern Ireland (30.1%) or Wales (36.1%). However, the proportion was significantly higher in Scotland than in England (25.3%) and the United Kingdom overall.

Across the other OECD countries included in the comparison, Scotland had a significantly lower proportion of pupils performing below Level 2 than 10 countries, including Iceland, Israel and Türkiye. Scotland was not significantly different from 20 countries, including Australia, Finland and France. Eight countries had a significantly lower proportion of pupils performing below Level 2 than Scotland, including Austria, Ireland and Switzerland.

At the upper end of the proficiency scale, 7.3% of pupils in Scotland performed at Level 5 or above. This was not significantly different from the OECD average, Northern Ireland (8.8%) or Wales (5.9%). However, the proportion of high-performing pupils was significantly lower in Scotland than in England (13.8%) and the United Kingdom overall.

Across the other OECD countries included in the comparison, 11 countries had a significantly higher proportion of pupils performing at Level 5 or above than Scotland, including Australia, Estonia, and Switzerland. Scotland was not significantly different from 14 countries, including Ireland, Sweden and Türkiye. Scotland had a significantly higher proportion of pupils performing at Level 5 or above than 12 countries, including France, Israel and Spain.

Overall, Scotland did not differ significantly from the OECD average in either the proportion of pupils performing below Level 2 or the proportion performing at Level 5 or above. Within the UK, Scotland did not differ significantly from Northern Ireland or Wales on either measure. However, compared with England and the UK overall, Scotland had a significantly higher proportion of pupils performing below Level 2 and a significantly lower proportion performing at Level 5 or above.

5.8 Mathematics performance by socio-economic background

PISA examines the relationship between pupils’ socio-economic background and mathematics performance using the PISA index of economic, social and cultural status (ESCS). This is based on information about pupils’ home and family circumstances, including parents’ education and occupations and resources available in the home.

In 2025, socio-economic background explained 7.8% of the variation in mathematics performance between pupils in Scotland. This means that differences in pupils’ socio-economic backgrounds accounted for a relatively small proportion of the overall differences in mathematics scores between pupils. This was significantly lower than in 2006, 2009, 2012 and 2022. It was not significantly different from 2015 or 2018. This indicates that socio-economic background was less strongly related to differences in pupils’ mathematics performance in 2025 than in a number of previous PISA cycles.

PISA also measures the socio-economic gradient, which shows how much mathematics performance tends to differ as pupils move up the ESCS scale. A steeper gradient indicates a stronger relationship between socio-economic advantage and performance, while a shallower gradient indicates a weaker relationship. In practical terms, this measure shows how much of a difference in mathematics performance is associated with being more socio-economically advantaged.

In 2025, a one-point increase on the ESCS scale was associated with an average increase of 27 points in mathematics performance. This means that pupils who were one point higher on the ESCS scale scored, on average, 27 points higher in mathematics. The gradient was significantly smaller in 2025 than in 2006, 2009, 2012 and 2022. It was not significantly different from 2015 or 2018.

Overall, pupils from more advantaged socio-economic backgrounds tended to perform better in mathematics in 2025. However, both measures indicate that the relationship between socio-economic background and mathematics performance was significantly weaker than in a number of previous PISA cycles, including 2022. This does not mean that socio-economic differences in performance have disappeared; rather, they were less strongly associated with mathematics performance than in those earlier cycles.

5.8.1 International comparisons of mathematics performance by socio-economic background

In 2025, socio-economic background explained 7.8% of the variation in mathematics performance in Scotland. This was significantly lower than in 21 countries, including Ireland, Australia, and Northern Ireland; and not significantly different from 18 countries, including England, Wales and Denmark. No country had a significantly lower proportion of variation explained by socio-economic background than Scotland. Scotland was also not significantly different from the United Kingdom as a whole and was significantly lower than the OECD average.

Scotland’s socio-economic gradient in mathematics was 27 points. The gradient was significantly smaller than in 22 countries, including Australia, Belgium and Northern Ireland; not significantly different from 16 countries, including England, Ireland and Wales, and significantly higher than one country, Mexico. Scotland’s gradient was also not significantly different from the United Kingdom as a whole and was significantly smaller than the OECD average.

Overall, both measures indicate that the relationship between socio-economic background and mathematics performance in Scotland was weaker than in many of the countries shown and weaker than the OECD average. However, Scotland did not differ significantly from England or Wales on either measure.

5.9 Mathematics performance by immigration background

PISA also examines mathematics performance by pupils’ immigration background. Pupils are classified as non-immigrant, second-generation immigrant (born in the UK but whose parents born elsewhere), and first-generation immigrant (born outside the country and whose parents were also born outside the country).

In 2025, non-immigrant pupils in Scotland achieved a mean mathematics score of 466 points. Second-generation immigrant pupils achieved a mean score of 504 points, which was significantly higher than the score for non-immigrant pupils. First-generation immigrant pupils achieved a mean score of 479 points, which was not significantly different from either non-immigrant or second-generation immigrant pupils.

Among non-immigrant pupils, mathematics performance in 2025 was significantly lower than in every assessment cycle between 2006 and 2018, but was not significantly different from 2022.

Among second-generation immigrant pupils, performance in 2025 was significantly lower than in 2012, but was not significantly different from 2006, 2009, 2015, 2018 or 2022. Among first-generation immigrant pupils, performance in 2025 was also significantly lower than in 2012, but was not significantly different from any of the other previous cycles.