PISA 2025 Results (Volume I): Luxembourg

The Programme for International Student Assessment (PISA) assesses the knowledge, skills and attitudes of 15-year-old students. The tests explore how well students can solve complex problems, think critically and communicate effectively; data collected through questionnaires describe the context in which students learn. This gives insights into how well education systems are preparing students for real life challenges and future success. Luxembourg has participated in PISA since 2000. By comparing results internationally, policy makers and educators in Luxembourg can learn from other countries’ policies and practices.

• Average 2025 results in Luxembourg were lower than in 2018 in mathematics and reading, and about the same as in 2018 in science.

• In science too, however, as in the other subjects, PISA 2025 results are below those observed a decade earlier, in PISA 2015.

• Over the most recent period (2018 to 2025), the gap between the highest-scoring students (10% with the highest scores) and the weakest students (10% with the lowest scores) did not change significantly in mathematics and reading, while it widened in science.

• Compared to 2015 the proportion of low-performing students (scoring below proficiency Level 2) increased by 11 percentage points in mathematics; by 13 percentage points in reading; and by six percentage points in science.

• Students in Luxembourg scored less than the OECD average in mathematics, reading and science (Figure 3).

• A similar number of students in Luxembourg to the average across OECD countries were top performers (Level 5 or 6) in at least one subject. At the same time, a smaller proportion of students than on average across OECD countries achieved a minimum level of proficiency (Level 2 or higher) in all three subjects (Figure 4).

• In computational problem solving, the mean score of students in Luxembourg was higher than the OECD average (Figure 3).

• Some 68% of students in Luxembourg attained Level 2 or higher in science (OECD average: 74%). At a minimum, these students can recognize the correct explanation for familiar scientific phenomena and can use such knowledge to identify, in simple cases, whether a conclusion is valid based on the data provided. Over 85% of students in top performers B-S-J-Z (China), Macao (China), Singapore, Estonia, Japan, Chinese Taipei and Korea (in descending order of that share) performed at this level or above.

• In Luxembourg, 9% of students were top performers in science, meaning that they were proficient at Level 5 or 6 (OECD average: 7%). These students can creatively and autonomously apply their knowledge of and about science to a wide variety of situations, including unfamiliar ones.

• In Luxembourg, 63% of students attained at least Level 2 proficiency in mathematics (OECD average: 65%). At a minimum, these students can interpret and recognize, without direct instructions, how a simple situation can be represented mathematically (e.g. comparing the total distance across two alternative routes, or converting prices into a different currency).

• Some 7% of students in Luxembourg were top performers in mathematics, meaning that they attained Level 5 or 6 in the PISA mathematics test (OECD average: 8%). The largest shares of students attaining Level 5 or 6 were observed in Seven Asian countries and economies: B-S-J-Z (China) (54%), Singapore (37%), Chinese Taipei (32%), Macao (China) (29%), Korea (23%), Japan (22%) and Hong Kong (China) (21%). At these levels, students can model complex situations mathematically, and can select, compare and evaluate appropriate problem-solving strategies for dealing with them. In 13 countries and economies participating in PISA 2025, more than 10% of students attained Level 5 or 6 proficiency.

• Some 62% of students in Luxembourg attained Level 2 or higher in reading (OECD average: 69%). At a minimum, these students can identify the main idea in a text of moderate length, find information based on explicit, though sometimes complex criteria, and can reflect on the purpose and form of texts when explicitly directed to do so. The share of 15-year-old students who attained minimum levels of proficiency in reading (Level 2 or higher) varied from 93% in B-S-J-Z (China) to 4% in Rwanda.

• In Luxembourg, 4% of students scored at Level 5 or higher in reading (OECD average: 6%). These students can comprehend lengthy texts, deal with concepts that are abstract or counterintuitive, and establish distinctions between fact and opinion, based on implicit cues pertaining to the content or source of the information.

• The PISA index of economic, social and cultural status is computed in such a way that all students taking the PISA test, regardless of the country where they live, can be placed on the same socio-economic scale. This means that it is possible to use this index to compare the performance of students of similar socio-economic background in different countries. In Luxembourg, 51% of students (the largest share) were in the top international quartile of the socio-economic scale, meaning that they were among the most advantaged students who took the PISA test in 2025. Their average score in science was 532 score points (Figure 5, left chart); one of the highest for students of similar socio-economic background.

• The PISA index of economic, social and cultural status can also be used to order students from the most disadvantaged to the most advantaged within each country and economy, and to create four groups of students of equal size (each comprising 25% of the population of 15-year-old students in each country/economy). In Luxembourg socio-economically advantaged students (the top 25% in terms of socio-economic status) outperformed disadvantaged students (the bottom 25%) by 130 score points in science. This is larger than the average difference between the two groups (85 score points) across OECD countries.

• Between 2015 and 2025, the gap in science performance between the top and the bottom 25% of students in terms of socio-economic status remained stable in Luxembourg, while the average gap across OECD countries narrowed. Over this period, performance remained stable both among advantaged and disadvantaged students.

• Socio-economic status was a predictor of performance in science in all PISA-participating countries and economies. It accounted for 22% of the variation in science performance in PISA 2025 in Luxembourg (compared to 12% on average across OECD countries).

• Some 7% of disadvantaged students in Luxembourg were able to score in the top quarter of science performance. These students can be considered academically resilient because, despite their socio-economic disadvantage, they have attained educational excellence by comparison with students in their own country. On average across OECD countries, 12% of disadvantaged students scored in the top quarter of science performance in their own countries.

• In Luxembourg, 71% of students agree or strongly agree that they are curious about many different things, compared to 73% across OECD countries. Moreover, 52% of students reported that they apply additional effort when work becomes challenging (OECD average 60%). At the same time, 29% agree or strongly agree that school has been a waste of time (OECD average: 24%).

• Curiosity, perseverance and valuing school are most often associated with higher science performance. In Luxembourg, after accounting for students’ and schools’ socio-economic profile, a one-unit increase in the curiosity index is associated with a difference of 24 score points in science performance, compared to an average of 19 score points across OECD countries; a one-unit increase in the perseverance index is associated with a difference of 17 score points in science performance (OECD average: 13 score points); students who perceive greater value in school and disagree that it is a waste of time tend to achieve higher scores in science (by 24 score points ) than their peers who feel they are wasting their time at school (OECD average: 27 score points).

• PISA 2025 also asked students about other attitudes towards learning. For example, 75% of students in Luxembourg have a growth mindset (OECD average: 69%), i.e. hold the belief that intelligence can be cultivated through work, effort and feedback. On average across OECD countries, students with a growth mindset score 28 score points higher in science than those with a fixed mindset. In Luxembourg, this difference is 23 score points, after accounting for students’ and schools’ socio-economic profile.

• PISA 2025 data show that there is still room for improvement in terms of students’ self-regulation and goal-setting behaviours. For example, only about 37% of students in Luxembourg report that they often or very often plan their approach to studying. Furthermore, 51% of students ask themselves questions about the material to check if they have understood everything (OECD averages: 37% and 49%, respectively).

• An autonomous self-regulated learner knows when and where to ask for help. PISA 2025 data show that, in most countries, students tend to rely on their teachers and classmates for support with their schoolwork. In Luxembourg, 67% of students ask their teachers for help when they need it and 79% ask their classmates when they do not fully understand something (OECD averages: 77% and 82%, respectively).

• In Luxembourg self-regulated goal-setting and help-seeking are positively related to science performance, after accounting for students’ and schools’ socio-economic profile. A one-unit increase in the index of goal-setting is associated with a difference of 4 score points (OECD average: 3 score points).A one-unit increase in the index of help-seeking is associated with a difference of 6 score points (OECD average: 6 score points).

• In 2025, 45% of students in Luxembourg were in schools whose principals reported that instruction was hindered, at least to some extent, by a perceived lack of teaching staff (OECD average: 39%); the corresponding share for a perceived lack of assisting staff was 36% (OECD average: 39%).

• PISA 2025 data reveal that perceived shortages of material resources remain a concern in most countries and economies. Some 1% of students in Luxembourg attended schools whose principals reported that a lack of educational material (e.g. textbooks, IT equipment, library or laboratory material) hindered instruction, at least to some extent, and 33% were in schools where a lack of physical infrastructure (e.g. buildings, grounds, heating/cooling systems, lighting and acoustic systems) was perceived to hinder instruction.

• In Luxembourg, students reported spending on average 1.6 hours per day using digital devices at school for learning activities (OECD average: 1.7 hours). At the same time, a non-negligible share of students’ device use on school days is dedicated to non-learning activities: students in Luxembourg reported spending 0.8 hours per day using digital devices at school for leisure (OECD average: 1.1 hours).

• In Luxembourg, 24% of students reported that their peers are often distracted by the use of digital devices during most or all science lessons (OECD average: 28%). On average across OECD countries, students who reported digital distraction in their science classes score 11 points lower in science than those who reported no digital distraction (after accounting for students’ and schools’ socio-economic profile). In Luxembourg they score 14 points lower.

• On average across OECD countries, students in schools where cell phone use is not allowed on school premises are about 13% less likely to report digital distraction. In 2025, 35% of students in Luxembourg attended schools with such a ban (OECD average: 49%).

• Students in most education systems use AI chatbots for schoolwork, creating new challenges and opportunities for their learning. In Luxembourg, 47% of students reported using AI chatbots to help them learn at least once a week (OECD average: 46%). Students also frequently use AI to conduct preliminary research on a new topic (Luxembourg: 34%; OECD average: 31%), to summarise a text they had to read (Luxembourg: 29%; OECD average: 30%), and to draft text for writing assignments (Luxembourg: 28%; OECD average: 29%).Only 11% of students reported never or almost never using AI chatbots for any of the schoolwork tasks examined in PISA (OECD average: 14%).

• In Luxembourg, 20% of students reported being victims of at least one form of bullying a few times per month or more in 2025 (OECD average: 20%).

• Cyberbullying is relatively uncommon compared with other forms of bullying, but its victims tend to experience significant achievement deficits. In Luxembourg, 4% of students reported that upsetting information about them was published online, without their consent, at least a few times a month during the 12 months prior to the PISA test (OECD average: 3%). These students scored, on average, 41 points lower in science than their peers.

• In Luxembourg, 58% of students reported that, in most science lessons, the teacher shows an interest in every student’s learning (OECD average: 69%), 58% that the teacher continues teaching until the students understand (OECD average: 66%), 67% that the teacher gives extra help when students need it (OECD average: 73%). In 2015, the corresponding shares were 57%, 63% and 67%.

• Disciplinary climate in science lessons in 2025 remains a concern in many countries and economies: about 20% of students in Luxembourg reported that they cannot work well in most or all lessons (OECD average: 19%); 33% reported noise and disorder disrupted most or every science lesson (OECD average: 31%); 33% of students reported that their classmates do not listen to what the teacher says (OECD average: 29%).

• Disciplinary climate and teacher support in science lessons improved between 2015 and 2025 on average across OECD countries. This was also the case in Luxembourg. Countries and economies with an improved disciplinary climate tend to show stability or an increase in their average science performance compared to 2015.

• In 2025, 67% of students in Luxembourg reported that their parents or guardians asked them, at least once per week, what they did at school , while 36% reported having at least weekly discussions about problems they might have at school.

• In 2025, around 15% of students in Luxembourg reported having skipped at least one day of school in the two weeks prior to the test (OECD average: 22%); 18% reported skipping at least one class (OECD average: 24%); 57% reported arriving late for school (OECD average: 50%). In previous years, the corresponding shares were 15%, 19% and 55% (in 2018).

• In Luxembourg, 67% of students reached at least baseline proficiency (Level 2) in environmental science. On average across OECD countries, 74% of students achieve this level, with more than 85% reaching it in seven countries and economies.

• PISA 2025 shows that students in most education systems report strong pro-environmental attitudes, although fewer feel confident about how to collaborate effectively. In Luxembourg, 75% of students believe that they can positively impact the environment (OECD average: 81%); 78% believe that collective action can address environmental problems (OECD average: 82%); 59% report knowing how to work with others to positively impact the environment (OECD average: 68%).

• In Luxembourg, 60% of students reported that they are somewhat or very interested in contributing to protecting the environment in their future jobs (OECD average: 62%).

• Most students engaged in environment-related activities at least once over the 12 months prior to the PISA test. On average across OECD countries, the most common forms of engagement are discussing with friends and family how to have a positive impact on the environment (Luxembourg: 60%; OECD average: 56%) and making purchasing decisions based on the sustainability of products (OECD average: 54%; 60% in Luxembourg).

• The PISA 2025 survey focused on science, with mathematics and reading as minor areas; an assessment of learning in the digital world measured students’ computational problem-solving skills and their ability to self-direct their learning in a digital environment. Results for science, mathematics, reading and computational problem solving are released on 8 September 2026 and analyses of students’ self-directed learning in 2027.

• Over 760 000 students took the assessment in 2025, representing about 33 million 15-year-olds in the schools of the 91 participating countries and economies.

• In Luxembourg, 6 078 students, in 49 schools, completed the assessment, representing about 6 500 15-year-old students (an estimated 91% of the total population of 15-year-olds).

• PISA students are aged between 15 years 3 months and 16 years 2 months at the time of the assessment, and they have completed at least 6 years of formal schooling. In this note, they are referred to as 15-year-old students.

• In PISA 2025, 65% of students in Luxembourg had an immigrant background (OECD average: 16%), meaning that both of the student’s parents were born abroad. This is higher than the share in 2018 (55%). Furthermore, 78% of students in Luxembourg speak a language at home that is different from the language in which they took the PISA assessment (OECD average: 15%).

• Students took a two hour test, and each student was tested in two subjects. Different students were given different test questions and different combinations of subjects (e.g. science followed by reading, or mathematics followed by science, etc.). Test items were a mixture of multiple-choice questions and questions requiring students to construct their own responses.

• Students also answered a background questionnaire, which took about 35 minutes to complete. The questionnaire sought information about the students themselves, their attitudes, dispositions and beliefs, their homes, and their school and learning experiences. School principals completed a questionnaire about school management, organisation, and the learning environment.

• Some countries/economies also distributed additional questionnaires, to students, parents and/or teachers, to elicit more information. The findings from these optional questionnaires are not covered by this note.