Nearly two in five 15-year-olds in the Netherlands did not reach the baseline reading proficiency level in PISA 2025, according to the OECD’s newly released country results. Only 61% of Dutch students reached Level 2 or higher in reading, compared with an OECD average of 69%. The result places the Netherlands below the OECD average in reading even as Dutch students remain stronger than average in mathematics and close to average in science.
The 39% figure is best understood as a proficiency threshold, not a count of teenagers who literally cannot read. PISA Level 2 represents the point at which students can identify the main idea in a moderately long text, locate information using explicit criteria and reflect on a text’s purpose and form when prompted. Students below that line can still read, but they are more likely to struggle as texts become longer, less familiar or more demanding.
The change is part of a sustained Dutch decline. Reading performance fell again between 2022 and 2025. Across reading, mathematics and science, the 2025 results are among the lowest the Netherlands has recorded in PISA and are significantly below its 2018 scores. The OECD describes the country’s trajectory since 2015 as clearly negative.
The composition of performance has also shifted. Only 4% of Dutch students reached Level 5 or higher in reading, compared with 6% across the OECD. At those upper levels, students are expected to work with lengthy texts, abstract or counterintuitive ideas and subtle distinctions between fact and opinion. Since 2015, meanwhile, the share of Dutch students below Level 2 in reading has increased by about 20 percentage points.
PISA is not a school examination tied to one national curriculum. It samples 15-year-olds and asks how well they can apply knowledge and skills to problems that extend beyond memorising classroom material. That makes the assessment useful for international comparison, although it cannot by itself identify why a country’s performance rises or falls.
The Dutch result sits inside a wider pattern. PISA 2025 covered 91 countries and economies, and the OECD reported that average reading and mathematics performance across its member countries has fallen to historically low levels. Long-term reading declines are visible in many systems, meaning the Netherlands is not an isolated case. Its 61% baseline attainment rate, however, remains below the OECD average.
Neighbouring Flanders offers a useful comparison. Nearly three in ten Flemish students fail to reach the reading baseline, according to the Flemish government, and its mean reading score has dropped sharply since 2015. The decline therefore crosses national borders in Dutch-language education, though its magnitude and distribution differ between systems.
The practical stakes extend beyond literature classes. Reading proficiency affects how students navigate science instructions, workplace procedures, health information, financial documents and digital media. In an information environment where text is abundant but source quality varies widely, comprehension and the ability to distinguish claims, evidence and opinion become part of basic civic and professional functioning.
The Netherlands now faces two linked questions. One is how to improve average reading performance. The other is how to reduce the size of the group below the minimum threshold while also rebuilding the upper end of the distribution. PISA cannot prescribe the intervention, but its trend data show that the problem has persisted long enough to outlast a single cohort or policy cycle.
For the teenagers represented in the 2025 assessment, the issue is immediate rather than theoretical. Many will move into vocational education, university preparation or work within the next few years. The next PISA cycle will measure whether the national trend changes, but this cohort will already be making decisions in settings that demand increasingly independent reading.
