Reference

What is a good IQ score?

The question has no answer until you say what for. Here are the six things people usually mean by it, and the threshold that applies to each.

1. Statistically above average: anything over 100

The narrowest reading. Anything above 100 is above the middle of the comparison group, and 110 puts you in the top quarter.

This is technically correct and rarely what people want, because it treats a one-point edge as meaningful. Measurement error alone is worth five points in either direction on a good clinical test, so 101 and 99 are the same result.

The useful version of this reading is the band rather than the point: 90 to 109 is typical, 110 to 119 is comfortably above typical, 120 and up is clearly above typical.

2. Enough for academic work: roughly 110 and up

Correlations between IQ and academic achievement run around 0.5, higher in earlier schooling and lower at postgraduate level where the range has already been narrowed by selection.

There is no threshold for university study, and there are large numbers of graduates across the whole average band. What the correlation implies is a probability shift, not a gate: a score around 110 makes academic work easier rather than possible, and effort and interest do a great deal of the remaining work.

At postgraduate level the observed correlation with performance drops sharply, which is a range-restriction artefact rather than evidence that ability stops mattering.

3. Enough for demanding professional work: usually quoted around 115–125

Occupational data consistently shows higher averages in professions requiring lengthy formal education and abstract problem-solving — medicine, law, research science, engineering.

The number that circulates is an average, not a requirement, and the difference matters. Within-occupation variation is consistently larger than between-occupation variation: the spread among accountants is wider than the gap between the average accountant and the average engineer.

So an occupational average tells you about the selection process that produced a profession, and almost nothing about whether a specific person can do the job. The occupational data in full.

4. Enough for Mensa: 130 on a standard scale

Mensa's requirement is the 98th percentile of the general population. On a scale with a standard deviation of 15 that is about 130.

The widely quoted figure of 140 is wrong, and the confusion comes from scales. On the Cattell scale, with a standard deviation of 24, the 98th percentile is around 148. Both describe the same rarity.

No online test qualifies you, including this one. Mensa admits members through its own supervised test or a prior result from a specific list of approved instruments. How entry works.

5. Enough for a gifted programme: usually 125–132

School gifted programmes typically use a cut-off around the 98th percentile, though some use 125 and some use a percentile on a group-administered test such as the CogAT instead.

The threshold is administrative rather than diagnostic. Nothing changes about a child between 129 and 131; programmes have a fixed number of places and a cut-off is a defensible way to allocate them.

Because a child's score carries a confidence interval of about plus or minus five points, a child whose true ability sits near the line will fall on either side of it depending on the day. Good identification practice uses multiple sources for exactly this reason. What parents should know.

6. Good for you: the version that actually matters

Here is the uncomfortable part. For almost everyone asking, the answer that would change anything is: it does not matter as much as you think.

IQ correlates with academic achievement at around 0.5 and with income at around 0.2 to 0.3. A correlation of 0.5 means that between two randomly chosen people, the higher scorer does better about 67 percent of the time — a real edge, and one that leaves a third of cases going the other way.

Conscientiousness predicts job performance about as well as cognitive ability in many roles and academic grades better than ability at university level. It is also the one of the two you can do something about.

If you came here after taking a test and you are trying to work out whether your number is good, the most accurate answer is that it places you in a band, the band is wide, the error on it is wider, and what you do next is more predictive than either.

The thresholds in one table