Which country has the highest average IQ?

Lists usually put East Asian places such as Singapore, Hong Kong, Japan, South Korea and Taiwan near the top, with averages around 105. The data behind these lists is uneven, so treat the order as approximate.

What is the average IQ in the United States?

Estimates for the United States usually fall between 97 and 100, which is close to the global reference score of 100.

Why can't I compare my score with a country average?

Tests are normed on their own sample. A score from one test is not directly comparable with a national figure built from other tests, years and sampling methods.

Does a country average say anything about an individual?

No. Differences within any country are far larger than differences between countries, so an average predicts almost nothing about one person.

Average IQ by Country: Read the Rankings Before You Quote Them

By Brain & IQ editorial team · Updated November 7, 2026

Dark globe covered with luminous violet and amber dots and lines

Type "average IQ by country" into a search engine and you get tidy tables with a flag next to each number. They look authoritative, they travel well on social media, and almost every one of them hides the same problem: nobody ever gave a single test to a fair sample of everyone in every country. The numbers are patched together. This guide shows you what is usually in those tables, why they differ, and how to use the information without embarrassing yourself in an argument.

Where the figures come from

Most country tables trace back to a compilation of studies published in the early 2000s and updated since, plus more recent estimates built from school assessments such as PISA and TIMSS. The first approach gathers individual IQ studies from each country, often small ones, sometimes from a single city, and averages them. The second treats results in maths and science as a stand-in for general cognitive ability, which is reasonable but not the same thing.

Both methods share three weaknesses:

Uneven samples. A country represented by a study of schoolchildren in the capital is not the same as a country measured nationally.
Different tests, different years. A Raven's matrices study from 1995 and a Wechsler study from 2018 are not on the same ruler, even if both report "an IQ".
Estimates filling gaps. Where no study existed, the missing country was sometimes given a number based on its neighbors. Those are guesses, however carefully made.

Because of this, serious researchers describe national IQ tables as rough indicators, not as precise facts, and they regularly disagree about methods.

What the lists usually say

With those caveats in mind, the broad pattern in nearly every ranking is stable enough to describe:

East Asia (Singapore, Hong Kong, Japan, South Korea, Taiwan, China) tends to sit at the top, with averages usually quoted between about 100 and 108.
Much of Europe, North America and Oceania clusters near 95 to 102.
Latin America, parts of the Middle East and South Asia are commonly listed in the 80s and low 90s.
The lowest figures in the tables come from regions where the data is thinnest, schooling and health conditions have been hardest, and the criticism of the methods is loudest.

The United States generally lands between 97 and 100. That is why "the average IQ" is usually quoted as 100: the scale is anchored on a mostly Western reference population.

Why countries differ at all

When a gap shows up, the explanation with the strongest evidence is not about nationality; it is about conditions. The factors that research keeps returning to include:

Schooling. Each extra year of education raises measured scores by a few points on average, according to a large meta-analysis of natural experiments. Countries with more and better schooling score higher.
Nutrition and health. Childhood malnutrition, iodine deficiency and infectious disease all drag on cognitive development. Improvements in these have driven some of the biggest national gains.
Familiarity with tests. Someone who has done multiple-choice reasoning problems since age six has a real advantage over someone seeing a matrix puzzle for the first time, even if their underlying reasoning is identical.
The Flynn effect. In many countries raw scores rose by about three points per decade through most of the twentieth century. In some wealthy countries the rise has slowed or reversed in recent decades, so when a table was built matters.

These factors are changeable, which is why the most honest reading of the tables is that they measure environments as much as people.

The thing most people get wrong

Differences between individuals within a country are enormous compared with differences between country averages. The standard deviation inside any one country is about 15 points, while the gap between, say, a high-scoring and a middle-scoring country is a handful of points. In practical terms, any national average is compatible with nearly every score from 70 to 130 turning up in every country. Using a country average to guess about a person is a statistical mistake, and a rude one.

How to compare your own score fairly

If you came here to see where you stand, a national table is the wrong ruler. Here is a better method.

Use the norms of the test you took. Your score is meaningful relative to the group that test was calibrated on, such as adults in an English-speaking country.
Read it as a range. An online result is best treated as a point plus or minus several points.
Compare against percentiles, not countries. A 115 puts you roughly in the top sixth of the reference group. That statement is far more precise than "above the average of Norway".
Retest after training, not after a nap. Differences you can act on come from practice with the format, which we cover in the Raven matrices guide.

Cognivo IQ Test is a 40-question test with 25 matrix puzzles, number series, logic items and short memory tasks, and its score is placed against a bell curve centered at 100. You receive your result, a certificate and an answer breakdown. The trial costs $1.00 for 7 days and then renews every 28 days until you cancel; the renewal amount is shown at checkout before you pay.

Questions worth asking of any ranking you meet

Before you share a table, ask four things. Who collected the data, and in which years? Was it a national sample or a handful of schools? Was the same kind of test used in every country, or were different instruments mixed? And were any values estimated from neighbors rather than measured? If a source cannot answer those, the ranking is decoration. If it can, you still have to remember that the unit being measured is a population average under particular conditions, not anyone's potential. Asking these questions takes a minute and makes you the most reliable person at the table.

A realistic way to use national IQ data

There are two cases where the numbers are genuinely useful. The first is development research: when a country invests in nutrition and schooling, average scores on standardized tests tend to rise within a generation, which is a good way of tracking whether the investment worked. The second is general curiosity, as long as you hold the figures loosely.

For everything else, including arguments at dinner, the safest sentence is this: "National averages are rough, they mostly reflect schooling, health and test familiarity, and they say almost nothing about any single person." That sentence will survive a fact check. Most of the tables on social media will not.

If you want to know about a person, the useful question is not what country they are from but how they reason on a task they have never seen. That is precisely what a well-made test measures, and what the score chart helps you interpret once you have your result.