The data sources: what we're actually looking at
The most widely cited national IQ dataset is Richard Lynn and Tatu Vanhanen's compilation, published in IQ and the Wealth of Nations (2002) and updated through subsequent editions. Heiner Rindermann's work extended and partially corrected this dataset, and David Becker maintains a frequently updated open-source compilation at viewoniq.org that many researchers now prefer for its greater transparency.
These datasets do not represent true national IQ measurements conducted to scientific standards. Rather, they aggregate available cognitive test data — some from formal IQ tests, much from international academic assessments like PISA (Programme for International Student Assessment), TIMSS (Trends in International Mathematics and Science Study), and PIRLS. Where no data exists at all, scores are estimated by extrapolation from neighbouring countries or from correlated development indicators.
The result is a ranking useful for identifying broad development-related patterns, but unreliable for precise country-to-country comparisons — especially for nations with limited or outdated data.
Top 25 countries by average IQ — ranked
The table below combines estimates from Lynn & Vanhanen (2012 edition), Becker's updated dataset (2023), and PISA 2022 results where available. The "reliability" column reflects data quality and representativeness of the underlying studies — it is not a judgement about the population being measured.
| # | Country | Est. Avg IQ | Score | Primary data source | Data reliability |
|---|---|---|---|---|---|
| 1 | 🇯🇵 Japan | 106–108 | Large representative samples + PISA top 3 | High | |
| 2 | 🇸🇬 Singapore | 105–108 | PISA & TIMSS top performer globally | High | |
| 3 | 🇹🇼 Taiwan | 104–106 | Multiple studies + TIMSS top tier | High | |
| 4 | 🇭🇰 Hong Kong | 103–108 | Multiple studies + PISA top 5 | High | |
| 5 | 🇰🇷 South Korea | 102–106 | Multiple studies + PISA top tier | High | |
| 6 | 🇨🇳 China | 102–105 | Urban samples only — rural data limited | Medium | |
| 7 | 🇫🇮 Finland | 101–103 | Multiple studies + PISA historically top 5 | High | |
| 8 | 🇨🇭 Switzerland | 100–102 | Multiple studies | High | |
| 9 | 🇳🇱 Netherlands | 100–102 | Multiple studies — strong data quality | High | |
| 10 | 🇩🇪 Germany | 99–102 | Multiple studies | High | |
| 11 | 🇦🇹 Austria | 99–101 | Multiple studies | High | |
| 12 | 🇬🇧 United Kingdom | 99–101 | Reference population for many IQ scales | High | |
| 13 | 🇸🇪 Sweden | 99–101 | Multiple studies + military conscript data | High | |
| 14 | 🇳🇿 New Zealand | 98–100 | Multiple studies including Dunedin cohort | High | |
| 15 | 🇺🇸 United States | 97–100 | Very large nationally representative samples | High | |
| 16 | 🇦🇺 Australia | 97–99 | Multiple studies | High | |
| 17 | 🇨🇦 Canada | 97–99 | Multiple studies | High | |
| 18 | 🇫🇷 France | 96–99 | Multiple studies | High | |
| 19 | 🇮🇱 Israel | 95–97 | Military conscript + ethnic sub-group samples | Medium | |
| 20 | 🇵🇱 Poland | 95–97 | Multiple studies — improving post-1990 | Medium | |
| 21 | 🇷🇺 Russia | 95–97 | Limited post-Soviet era studies | Medium | |
| 22 | 🇧🇷 Brazil | 83–87 | Multiple studies — improving trend | Medium | |
| 23 | 🇲🇽 Mexico | 82–88 | Limited samples — wide range across studies | Low | |
| 24 | 🇮🇳 India | 76–82 | Mixed samples, extreme regional variation | Low | |
| 25 | 🇳🇬 Nigeria | 68–75 | Very limited samples — highly contested estimates | Low |
These scores should not be read as fixed characteristics. The same methodology that gives Japan a score of 106 today would have given Japan a score of around 78–80 using 1950 Western norms. The data reflects current environments, not inherent population differences — and environments change.
The Flynn Effect: IQ scores are rising everywhere
The single most important fact for interpreting any national IQ comparison is the Flynn Effect, named after researcher James Flynn who documented it systematically in 1987. Average IQ scores have risen substantially — by 3 IQ points per decade in many industrialised nations, and by considerably more in some developing countries undergoing rapid environmental improvement.
The magnitude of this effect is staggering. A person of average intelligence in 1900, tested by modern standards, would score approximately 70 on today's IQ scale — a score we would now classify as borderline intellectual disability. Conversely, the typical person today would have appeared exceptionally gifted by 1900 standards. Yet the human gene pool has not changed meaningfully in 100 years. The entire shift is environmental.
The primary drivers identified in the research literature are: expansion of formal education (which directly trains abstract reasoning), improvements in childhood nutrition — particularly iodine supplementation programmes, which alone account for several IQ points in countries where deficiency was prevalent — reduction in infectious disease burden, removal of environmental neurotoxins such as lead from fuel and paint, and greater exposure to visually complex and abstract media.
Flynn Effect gains by region
Flynn gains have not been uniform globally. They track closely with patterns of environmental improvement — following education access, nutritional programmes, healthcare expansion, and economic development.
| Region | Approx. gain / decade | Period | Primary environmental drivers | Current status |
|---|---|---|---|---|
| Western Europe & Scandinavia | +3 pts | 1950–2000 | Education quality, nutrition, healthcare | Plateauing / slight reversal from ~2000 |
| United States | +3 pts | 1930–2010 | Education expansion, lead removal, nutrition | Slowing significantly post-2000 |
| East Asia (Japan, South Korea, Taiwan) | +7–8 pts | 1950–2000 | Rapid education expansion, urbanisation, nutrition improvements post-war | Largely plateaued — near ceiling |
| Latin America | +4–6 pts | 1970–2020 | Education access, nutrition programmes, urbanisation | Ongoing — gains continuing |
| Sub-Saharan Africa | +3–5 pts | 1980–2020 | Education expansion, malaria & disease burden reduction | Ongoing — large gains still possible |
| South & Southeast Asia | +4–6 pts | 1970–2020 | Education, urbanisation, iodine supplementation | Ongoing — accelerating in some regions |
The East Asian trajectory is the most instructive case in the entire dataset. Japan, South Korea, and Taiwan had average IQ scores estimated 20–25 points below contemporaneous Western norms in 1950. Within two generations — through post-war reconstruction, universal high-quality education, and nutritional improvements — they reached the top of global rankings. No genetic explanation can account for a change of that speed and magnitude. This is what environmental investment in human development actually looks like in the data.
What drives national IQ differences
The research literature identifies a consistent set of environmental factors that predict national average IQ scores. For most of them, there is strong causal evidence from intervention studies and natural experiments — not just correlation.
Formal education quality and duration. Every additional year of education is associated with an IQ increase of 1–5 points in meta-analyses (Ritchie & Tucker-Drob, 2018, covering 615,000 participants). The effect is particularly strong for abstract and fluid reasoning — the domains most directly trained by structured schooling. Countries with universal, high-quality compulsory education score substantially higher than countries with similar economic development but weaker school systems.
Childhood nutrition — particularly iodine and iron. Iodine deficiency is the world's leading preventable cause of brain damage and is strongly associated with lower IQ at the population level. Studies of iodine supplementation programmes in deficient regions show IQ gains of 10–15 points in affected populations (Bleichrodt & Born, 1994). Iron deficiency anaemia in early childhood similarly impairs cognitive development. Many low-scoring countries are also high-burden iodine and iron deficiency countries.
Infectious disease burden. The parasite stress hypothesis, developed by Eppig, Fincher & Thornhill (2010), proposes that cognitive resources normally available for learning and brain development are redirected to immune function in high-disease environments. Their cross-national analysis found that parasite prevalence was the strongest single predictor of national IQ — stronger than GDP or education. This is a causal claim consistent with the biological mechanism and with historical data on disease eradication and IQ gains.
Lead and neurotoxin exposure. The phasing out of leaded petrol in the United States between 1970 and 1990 is estimated to have raised average IQ by 2–5 points and reduced the proportion of children with blood lead levels above the threshold for cognitive impairment dramatically. Lead particularly damages the developing prefrontal cortex. Countries that industrialised rapidly with weak environmental regulations show higher historical neurotoxin burdens — and lower IQ estimates.
Healthcare and maternal health. Prenatal care quality, rates of birth complications, and postnatal nutrition all affect early brain development. Countries with strong universal healthcare systems — particularly in prenatal and early childhood — show better cognitive outcomes controlling for income.
Methodological problems with the data
Acknowledging the limitations of this data is not an optional disclaimer — it is essential for any responsible use of the rankings.
| Problem | What it means in practice | Countries most affected | Direction of bias |
|---|---|---|---|
| Small unrepresentative samples | Some national estimates derive from a single study of 50–300 students, often urban school pupils — not representative of the national population including rural areas | Sub-Saharan Africa, Central Asia, Pacific islands | Likely underestimates (urban students score higher than rural average) |
| Outdated data | Many estimates use studies from the 1970s–90s in countries that have developed substantially since. Flynn gains will have raised actual current scores. | South/Southeast Asia, Latin America, parts of Africa | Systematic underestimation of current scores |
| Non-equivalent tests across countries | Different tests measure different cognitive constructs to different degrees. A Raven's Progressive Matrices score cannot be directly equated to a WAIS score. | All countries — magnitude varies by test used | Adds noise; direction depends on which test was used |
| Test motivation & familiarity | Populations with less experience of standardised testing often perform below actual ability. Effort levels and test-taking strategies vary culturally. | Countries with limited standardised testing history | Systematic underestimation in less-tested populations |
| Imputed scores — no actual data | Many countries have zero IQ test data. Their scores are extrapolated from academic assessments, neighbouring country data, or development indicators. | Much of Central Africa, Central Asia, Pacific | Highly uncertain; methodology is scientifically disputed |
| China's sampling bias | Chinese samples are almost exclusively urban and coastal. Rural inland China (40%+ of the population, with lower education access) is severely underrepresented. | China | Likely significant overestimate of national average |
What you should not conclude from this data
National IQ averages are among the most frequently misused data points in public discourse. The following conclusions are not supported by the evidence:
That national differences are genetic. The Flynn Effect is definitive evidence against genetic determinism for current national IQ differences. A country's average can change by 15–30 points within a generation through environmental change alone. Ireland scored in the low 80s by historical estimates in the early 20th century; by the 1990s it had reached the European average through educational investment. The genetic composition of Ireland did not change in 50 years.
That individuals from low-scoring countries are less capable. Population averages tell you nothing about individuals. The IQ score distribution in every country has enormous variance — the highest-scoring 10% of individuals in Nigeria score well above the median in Japan. Applying group statistics to individuals is not just morally wrong — it is a fundamental statistical error.
That rankings are precise. The range of estimates for most countries spans 5–10 IQ points across different studies. Treating any country's score as a specific single number is false precision. The China estimate alone spans from 98 to 108 across credible studies.
That low scores reflect fixed potential. They reflect current conditions. Every historical case of rapid development — East Asia post-1950, Ireland post-1970, the Netherlands recovering from the 1944 famine — shows large and rapid IQ gains following environmental improvement. Low scores indicate developmental opportunity, not permanent limitation.
What the data honestly does support
Used carefully, national cognitive data provides genuine insights into global development patterns. There are measurable real differences in average cognitive test performance across countries — these are not artefacts of measurement bias, though bias does affect their magnitude.
These differences correlate strongly with development indicators: years of schooling, childhood nutrition, healthcare access, income inequality, and disease burden. The causal direction predominantly runs from environmental conditions to cognitive scores — which means improving those conditions is the lever for raising scores, not the reverse.
The data makes the strongest possible argument for investment in education quality, nutritional programmes, and healthcare access in developing countries. Not as charity, but as the most reliable mechanism for raising human cognitive capability at scale. Every country that has invested heavily in these conditions has seen exactly the IQ gains the environmental model predicts.
For a deeper understanding of what IQ tests actually measure and what scores mean at the individual level, see our article What Is IQ?. For how scores change across the human lifespan, see Average IQ by Age. If you want to test yourself using a properly scored adaptive test, the AurorIQ test is free and requires no email signup.
Frequently asked questions
Which country has the highest average IQ?
Japan, Singapore, Taiwan, Hong Kong, and South Korea consistently appear at the top of national IQ rankings, with estimated average IQs of 104–108. These scores are supported by strong performance on international academic assessments (PISA and TIMSS) and are associated with high investment in education, good childhood nutrition, and low disease burden. The data for these countries is among the most robust in the national IQ literature.
What is the average IQ of the United States?
The United States average IQ is estimated at 97–100, placing it among high-scoring nations but below several European and East Asian countries. US data is among the most robust in the literature — there are multiple large nationally representative samples including the WAIS standardisation studies. The US score is pulled somewhat below the UK reference average in some estimates due to educational inequality across states and school districts.
Why do IQ scores differ between countries?
National IQ differences are primarily driven by environmental factors: quality and duration of formal education, childhood nutrition (especially iodine and iron intake), access to healthcare, infectious disease burden, and exposure to neurotoxins like lead. The Flynn Effect — IQ gains of 3–8 points per decade in many countries — proves that environmental improvements directly raise scores. Countries that were 20–30 points below Western norms in 1950 now score comparably after environmental development. Genetic explanations for current national differences are not supported by this evidence.
Is average IQ by country data reliable?
Reliability varies enormously. For high-income nations with long research traditions — Japan, US, UK, Germany, Netherlands, Sweden — estimates are based on multiple large representative studies and are reasonably reliable. For many developing countries, estimates are based on small, non-representative samples or extrapolated from academic assessments. Many countries in sub-Saharan Africa, Central Asia, and the Pacific have no direct IQ test data — their scores are imputed. Low-reliability estimates should be treated with significant scepticism.
What is the Flynn Effect and why does it matter?
The Flynn Effect is the documented rise in average IQ scores over the 20th century — approximately 3 points per decade in many industrialised nations, and more in some developing countries. Named after researcher James Flynn (1987), it demonstrates that IQ is highly responsive to environmental conditions. The rise is too rapid to have a genetic explanation — it reflects improvements in education, nutrition, disease control, and removal of neurotoxins. For national IQ comparisons, the Flynn Effect is the definitive evidence that current differences reflect current environmental inequality, not fixed genetic differences between populations.
Does a low national IQ mean people from that country are less intelligent?
No. A low national average reflects current environmental conditions — primarily educational access, nutrition, and healthcare — not fixed or inherent potential. IQ score distributions in every country overlap massively: the highest-scoring individuals in any lower-ranked country score well above the average in any higher-ranked country. Population averages cannot predict individual performance. Countries with previously low scores that subsequently improved their environments consistently show large gains — Ireland, Japan, South Korea, and Taiwan are all examples of this.