Average IQ
by country.

At a glance

What the country IQ table actually says

The short version, before the detail. Every figure below is taken from the same published dataset and can be checked against the full table further down this page.

The map

Average IQ by country, mapped

Darker blue is a higher published figure; orange is lower. Grey means the dataset has nothing for that territory at all. Hover or tap any country for its score and rank.

The top of the table

Countries with the highest average IQ

The leading group has barely changed across every version of this dataset since 2002: a tight cluster of East Asian economies, followed by northern and central Europe.

The top 25, ranked

Three things are worth noticing about this list before drawing any conclusion from it.

The margins are tiny. First place and tenth place are separated by under six points — less than half a standard deviation, and well inside the measurement error of most IQ tests. Treating a ranking built on those margins as a league table of national ability is not something the data supports.

The leaders share circumstances, not ancestry. The countries at the top have long compulsory schooling, near-universal literacy, low childhood disease burden and a culture in which formal testing is routine from an early age. Those are exactly the conditions known to raise measured scores.

The same countries lead PISA. Singapore, Japan, South Korea, Taiwan, Hong Kong and Macau occupy the top of the OECD's international student assessments as well. Two different instruments pointing the same way tells you the schooling effect is real; it does not tell you the cause is innate.

The full dataset

Average IQ by country: all 201

Search for a country, filter by region, or sort any column. Values are read directly from the published dataset without alteration — including the ones we think are wrong, which are discussed below. The Sample column is the part every other ranking leaves out: est. means no test data of any kind, and sch. means school tests rather than an IQ study.

The bottom of the table

The lowest figures — and what they measure

This is the part of the dataset that has drawn the heaviest criticism, and the part we would urge you to read most sceptically. We publish it because omitting it would be dishonest — not because we think these numbers mean what they appear to mean.

By region

Average IQ by region

Each figure below is the unweighted mean of the countries listed in that region — a simple average of national values, not a population-weighted one.

Provenance

Where these numbers actually come from

Almost every "average IQ by country" page on the internet — including the ones that rank above this one — is republishing the same single dataset. It is worth knowing what that dataset is.

The source is Richard Lynn and David Becker’s The Intelligence of Nations (2019), published by the Ulster Institute for Social Research, and the NIQ dataset released alongside it. The figures on this page are the QNW+SAS+GEO column of dataset version 1.3.2, read directly from Becker’s published workbook rather than from a secondary copy. It pools 669 studies covering 617,581 people and converts each onto a common scale anchored to a British mean of 100.

A quarter of the table was never tested

Of the 201 entries, only 130 rest on actual IQ studies. Another 18 are derived from school-assessment scores rather than an IQ test. The remaining 53 — 26% of the table — have no test data of any kind; their figure is an average of the neighbouring countries that do.

This is not a footnote. North Korea ranks sixth in the world on this table, at 103.76, and there is no IQ study of North Korea in the dataset at all — the number is inferred from its neighbours. Liechtenstein, tenth at 101.07, has no IQ study either. Two of the thirteen countries scoring above 100 got there without anyone being tested.

You can see the mechanism in the table above without taking our word for it. Search for Fiji: it sits on 83.96, and so do the Solomon Islands, Samoa, Kiribati, Micronesia, Tonga, the Marshall Islands and the Cook Islands — identical to two decimal places. That is not eight findings. It is one estimate printed eight times.

Where that estimating step can be checked against real school testing, it does not work: for the countries whose value is purely geographic guesswork, the imputed figure correlates r = −0.376 with independently derived achievement scores — a negative relationship.

The samples were never designed to represent nations

A nationally representative cognitive survey is expensive, and very few countries have ever run one. What exists instead is a patchwork: school classes, army recruits, university volunteers, clinical comparison groups. The United States entry pools 62,649 people across 58 studies. Angola’s rests on 19 people in a single study; the Dominican Republic’s on 34; Uzbekistan’s on 51; Namibia’s on 103. Every one of those figures is printed to two decimal places, exactly like the American one. That is why we show the sample size beside every country.

The studies span seventy years, and scores moved over that time

Measured IQ rose substantially across the twentieth century — the Flynn effect. A study run in 1975 and a study run in 2015 are not on the same footing unless that drift is corrected for, and countries are not all represented by studies from the same era. Some of the gaps in this table are gaps in time as much as gaps in ability.

The criticism

How reliable is national IQ data?

Short answer: reliable enough to be interesting, nowhere near reliable enough to rank populations by intelligence. Here is the case against it, made properly rather than waved away.

The drivers

Why average scores differ between countries

Measured differences between countries are real as measurements. What produces them is much better understood than the rankings suggest — and almost all of it is environmental and changeable.

Our own explainer on the factors that affect IQ test results covers how several of these operate at the level of a single test sitting rather than a whole country.

Change over time

The Flynn effect — and the reversal

Measured IQ is not fixed at the level of a population either. It rose for most of the twentieth century, nearly everywhere it was measured, and in some rich countries it has since started falling again.

Scores rose by roughly three points per decade through the twentieth century — around thirty points in total. A person scoring exactly at the 1930s average would land near the second percentile against today’s norms. The meta-analytic estimate across 285 studies is 2.31 points per decade, rising to 2.93 on modern Wechsler and Stanford-Binet tests. Nothing about human biology changed over those hundred years. Schooling, nutrition, family size and the sheer amount of abstract symbolic work in ordinary life all did.

This matters directly for the table above: a country represented by a study from 1975 and a country represented by one from 2015 are not being compared on equal terms. Some of the spread in any national IQ ranking is a gap in when people were tested rather than in how they scored.

And in several rich countries it has reversed

The same direction shows in international schooling data: average OECD performance in PISA fell 15 points in mathematics and 10 in reading between 2018 and 2022, the sharpest drop since the programme began. Whether that is a lasting reversal or a pandemic scar is still argued. Either way it undercuts any reading of a national IQ table as a permanent ordering of peoples — the countries at the top are the ones currently moving down.

Correlations

What national IQ tracks

National IQ estimates are not random noise. They correlate strongly with other measures of cognitive performance — which is exactly why the environmental explanation is the compelling one.

Here is what they actually track, from the published literature. National IQ correlates with educational attainment at r = 0.92, with international student assessments such as PISA and TIMSS at 0.85, with mean years of schooling at 0.76, with adult literacy at 0.64 — and with child mortality at −0.84. Against log GDP per capita it manages 0.69.

Notice the ordering. The tightest relationships are all with schooling and child health; income comes some way behind. And when Wicherts, Borsboom and Dolan placed national IQ in a principal components analysis alongside 17 development and geography indicators, a single component absorbed 65% of the variance, with national IQ loading 0.873 on it — right beside secondary-school enrolment at 0.943, fertility at 0.910, child mortality at 0.908 and sanitation at 0.895. Statistically, the measure behaves like one more indicator of a country’s development level.

The PISA cross-check

The OECD’s PISA programme tests fifteen-year-olds in 81 education systems on a common instrument, with proper sampling — everything the national IQ dataset lacks. Its 2022 leaders were Singapore (575 in mathematics), Macau, Taiwan, Japan, South Korea and Hong Kong, with Estonia the highest-scoring non-Asian system and Canada, Ireland and Switzerland close behind. That is very nearly the top of the IQ table, arrived at independently.

The instructive case is Vietnam: 34th in PISA, well above what its income or its average schooling would predict, and 60th in the IQ table at 89.5 — noticeably higher than its neighbours and its GDP per capita would suggest. Countries that invest early in schooling outperform their wealth on both measures. That is a policy finding, not a genetic one.

The overlap

Why a country average says nothing about you

This is the single most misread thing on this page, so it is worth showing rather than asserting. Two national distributions ten points apart still overlap almost completely.

Reading a score

What counts as a good IQ score?

IQ is not a raw total. It is a position on a curve, built so that the population mean is 100 and the standard deviation is 15. That construction is what makes any of these comparisons possible.

We go through this in more depth in IQ score ranges explained, the IQ level chart and IQ classifications. If you are curious where the threshold for the top tier sits, what counts as a genius IQ level covers it, and what an average IQ test score means is the companion piece to this page.

Country spotlights

A closer look at ten countries

The same figure means slightly different things depending on how it was collected. These are the countries people search for most, with the context the ranking alone leaves out.

Questions & answers

Country IQ: common questions

Methodology

Sources & how we built this page

We publish the dataset in full, name it, and publish the criticism of it in the same place. Every figure on this page can be traced to one of the sources below.