School pupils in Wales are falling behind in reading according to the latest international data.
The 2025 Programme for International Student Assessment (Pisa) is run by Organisation for Economic Cooperation and Development (OECD) and compares the educational skills of 15-year-olds around the world every three years.
In 2025, 15-year-olds in Wales achieved an average reading score of 459, not hugely different from the OECD average of 461. But Wales remained significantly below England (497), Scotland (488) and Northern Ireland (482).
Being close to the OECD average does not mean reading in Wales has improved. Wales scored 466 in 2022 and 483 in 2018. Instead, the OECD trend average fell by 14 points between 2022 and 2025.
Almost one in three pupils in Wales (31%) did not reach Pisa’s baseline level two in reading, the level the OECD considers necessary for young people to participate meaningfully in society. There was also a striking gender difference. Girls scored 472 and boys 446, with the gap widening from 11 points in 2015 to 26 in 2025. This was driven by a significant decline among boys while girls’ performance remained statistically stable.
Reading underpins learning across the curriculum. But Pisa should be seen as a warning light, rather than a diagnosis of what is wrong with Welsh education.
What does Pisa tell us?
Pisa examines how well 15-year-olds can apply their knowledge and skills rather than recall a particular curriculum. Reading literacy means more than reading words accurately. Pupils must understand and evaluate texts, locate information and distinguish fact from opinion. Pisa therefore provides a valuable snapshot of how young people in Wales are performing and allows us to examine both international differences and how a country changes over time.
International comparisons also demonstrate what is possible. Ireland, for example, scored 500 in reading in 2025, compared with Wales’s 459, although Ireland itself experienced a significant fall since 2022. Such comparisons cannot tell Wales what to copy, but they show that substantially higher levels of reading attainment are achievable.
What Pisa cannot tell us
Pisa assessed 3,406 learners from 113 Welsh schools. Wales met the OECD’s response-rate standards, although precise comparisons require some caution. The more important limitation is that a national average cannot tell us why an individual child struggles to read.
Pisa is observational. It cannot establish that a particular curriculum, teaching method, poverty, bilingualism or use of digital technology caused differences between countries. Nor can the results themselves explain the gap between Wales and the other UK nations.
Reading is complex. Children need to understand how letters represent spoken sounds, but skilled reading also depends on oral language, vocabulary, fluency and comprehension.
Most importantly, Pisa cannot tell a teacher what a particular child needs to learn next.
That points to a more fundamental challenge. We should not wait until children are 15 to discover that they have difficulty reading. Wales needs good information about literacy development throughout primary and secondary education, so that difficulties can be identified early enough to act.
Wales has already begun to change direction
If Wales is to improve these outcomes, the problem needs to be identified much earlier. Good assessment should identify children’s strengths and difficulties, follow their development over time and help teachers decide what support is needed. Assessment used in this way is very different from testing schools for accountability.
And identifying difficulties matters only if effective support follows. There is good evidence that targeted intervention can work. In two recent randomised trials, our research group found that structured literacy programmes delivered by teaching assistants improved reading and vocabulary among struggling readers in both English- and Welsh-medium schools, with gains in reading comprehension in the English trial and some cross-language transfer – applying skills from one language in another – in the Welsh trial.
There has recently been a significant shift in Wales’s approach. Following concerns about attainment, the Welsh government established an independent Literacy Expert Panel, which has published national principles for teaching language and literacy and a new statement on early reading. These emphasise systematic, structured phonics teaching alongside the rich spoken language, vocabulary and comprehension experiences that children need to become skilled readers.
This is being accompanied by investment in a more coherent national literacy infrastructure: professional learning for teachers, evidence-informed teaching programmes and interventions, and assessments designed to identify children’s strengths and difficulties. CAL:ON Cymru, which we lead and which is funded by the Welsh government, is part of this work, developing bilingual literacy assessments, teaching programmes and professional learning for schools across Wales.
The pupils taking Pisa in 2025 began learning to read around a decade ago. These results therefore tell us considerably more about the education they experienced than about initiatives introduced in the last year or two.
The latest results should be taken seriously, but repeatedly measuring a problem will not solve it. The more important test is whether Wales can now identify reading difficulties earlier and give teachers the evidence, expertise and tools to act on them.
