Korea has participated in every PISA cycle from 2000 through 2025. In PISA 2025, students in Korea averaged 526 points in science, 522 in mathematics and 501 in reading. All three means were above the corresponding OECD averages, but the long-run paths differ sharply by subject.
Korea’s PISA 2025 Scores
Latest mean scores compared with OECD averages from the same assessment cycle.
The OECD reports that Korea performed above the OECD average in science, mathematics and reading in PISA 2025. The largest difference based on the published rounded means was in mathematics, where Korea’s mean was 59 points above the OECD average. These arithmetic gaps are descriptive comparisons calculated from the published means rather than separate OECD indicators.
Science, mathematics and reading are separate PISA scales. Adding or averaging Korea’s three domain scores would create a measure that the OECD does not report as an official overall PISA score.
How Korea’s Scores Changed From 2000 to 2025
Reading has a comparable series from 2000, mathematics from 2003 and science from 2006.
Korea’s PISA Score Trends by Domain
Published mean scores across Korea’s PISA participation history
The vertical scale runs from 480 to 570 points to make cycle-to-cycle movement readable without reducing the range to the observed minimum and maximum. Missing early values reflect the starting points used for comparable subject trends, not missing Korean participation.
Korea PISA Scores by Assessment Cycle
Mean PISA score points. An em dash marks cycles before the standard long-term comparison series for that subject.
| PISA Cycle | Mathematics | Reading | Science |
|---|---|---|---|
| 2000 | — | 525 | — |
| 2003 | 542 | 534 | — |
| 2006 | 547 | 556 | 522 |
| 2009 | 546 | 539 | 538 |
| 2012 | 554 | 536 | 538 |
| 2015 | 524 | 517 | 516 |
| 2018 | 526 | 514 | 519 |
| 2022 | 527 | 515 | 528 |
| 2025 | 522 | 501 | 526 |
Historical means are shown as rounded whole score points. The OECD’s statistical comparisons use underlying unrounded estimates and their sampling uncertainty.
- Reading has moved furthest from its series high. Korea recorded 556 points in 2006 and 501 in 2025, a 55-point difference between the published rounded means.
- Mathematics reached 554 points in 2012. The 2025 mean of 522 is 32 points below that historical high.
- Science has followed a different path. The highest rounded mean was 538 in both 2009 and 2012. Korea scored 526 in 2025, 12 points below that level and close to its 2022 result.
Long-Run Movement Depends on the Subject
Simple endpoint differences use each subject’s first standard comparison year and the latest published mean.
These endpoint differences are calculated from rounded published means. They are useful for describing where the series started and ended, but they are not substitutes for OECD trend estimates or statistical tests based on all available cycles.
The shape of the series matters. Mathematics did not fall steadily from 2003: Korea moved upward through 2012 before the mean dropped to 524 in 2015 and then remained in a narrower 522–527 range from 2015 through 2025. Reading reached its highest published mean earlier, in 2006, and later settled near 514–517 across the 2015, 2018 and 2022 cycles before moving to 501 in 2025.
Science shows less similarity to those two patterns. The mean rose from 522 in 2006 to 538 in 2009 and 2012, fell to 516 in 2015, then moved to 519 in 2018, 528 in 2022 and 526 in 2025.
The 2022–2025 Change Was Concentrated in Reading
OECD short-term change estimates use unrounded score estimates rather than subtraction of displayed whole numbers.
The OECD describes Korea’s 2025 results as about the same as 2022 in mathematics and science, while reading was lower than in 2022. The same distinction appears when the latest results are compared with roughly a decade earlier: the OECD describes mathematics and science as on par with 2015, while reading is lower.
The displayed reading means are rounded to whole points. Subtracting 501 from 515 gives 14, but the OECD reports a 15-point 2022–2025 change using the underlying unrounded estimates. The official change figure should therefore be used when describing the short-term movement.
What the 2025 Mean Scores Do Not Show
Students with the same country mean can be distributed differently across PISA proficiency levels.
| Domain | Below Level 2 | Levels 5–6 |
|---|---|---|
| Science | 14.2% | 14.7% |
| Mathematics | 18.6% | 23.2% |
| Reading | 17.8% | 10.3% |
Korea’s 2025 Proficiency Distribution by Domain
Share of students below Level 2, at Levels 2–4 and at Levels 5–6
The Levels 2–4 shares are calculated as 100% minus the official below-Level-2 and Levels-5–6 shares. Percentages may be affected by rounding.
Mathematics combines Korea’s largest share of top performers among the three domains, 23.2%, with an 18.6% share below Level 2. Science has the smallest below-Level-2 share, at 14.2%. Reading has a similar low-performer share to mathematics but a smaller Levels-5–6 group, at 10.3%.
A country mean summarizes performance across the tested student population. It does not mean that every student performs near that value, nor does the mean assign an individual proficiency level to a typical student.
Why the Three Historical Series Start in Different Years
Korea first participated in PISA in 2000 and has taken part in every subsequent cycle: 2000, 2003, 2006, 2009, 2012, 2015, 2018, 2022 and 2025. The interruption between 2018 and 2022 reflects the rescheduling of the assessment cycle rather than a missed Korean administration.
Participation history and subject trend history are not identical. The OECD uses 2000 as the standard starting point for long-term reading comparisons, 2003 for mathematics and 2006 for science. Those were the first cycles in which each subject served as the major assessment domain and established the scale used for the standard long-term series.
Data Sources and Trend Measurement
The main source for Korea’s 2025 science, mathematics and reading results, comparisons with the OECD average, proficiency information and the OECD’s interpretation of recent score changes.
Provides the official 2025 country and economy means, OECD averages, short-term changes and statistical comparison tables used for the latest assessment.
The PISA 2022 performance-trends tables provide Korea’s published historical mean scores through 2022 and document the different comparison starting years for reading, mathematics and science.
The OECD participant profile confirms Korea’s participation in every PISA cycle from 2000 through 2025.
OECD Education GPS provides the detailed 2025 shares used here for students below Level 2 and students at Levels 5–6 in Korea.
PISA mean scores are estimates based on samples of 15-year-old students and carry statistical uncertainty. Whole-number scores in historical tables are rounded. Small differences between cycles should not automatically be treated as statistically detectable changes. Where the OECD publishes a formal comparison, its unrounded estimates and uncertainty measures take precedence over subtraction of displayed rounded values.