IQ score ranges: a complete interpretation guide
IQ scores are built to follow a bell curve with an average of 100 and a standard deviation of 15. That construction — not any natural law — is what gives the familiar bands their shape. This guide translates each band into percentiles, explains the error bar nobody shows you, and sets out how to read a result without over-reading it.
The bands at a glance
| Score range | Share of people | Percentile |
|---|---|---|
| Below 70 | ~2% | Below the 2nd |
| 70 – 84 | ~14% | 2nd – 16th |
| 85 – 99 | ~34% | 16th – 50th |
| 100 – 114 | ~34% | 50th – 84th |
| 115 – 129 | ~14% | 84th – 98th |
| 130 and above | ~2% | 98th and above |
What each band actually means
Below 70
Historically one criterion in assessing intellectual disability, but a diagnosis always requires evaluation of adaptive functioning by a qualified professional — never a test score alone.
70 – 84
Below average on abstract reasoning tasks. Many people in this band study, work and live entirely ordinary lives; the score describes test performance, not competence.
85 – 99
Lower half of the average range. The distance between 88 and 98 is comparable to normal day-to-day variation within one person.
100 – 114
Upper half of the average range. 100 is the definition of the midpoint, not a pass mark.
115 – 129
Above average. Roughly the range where dense, abstract material becomes comfortable rather than effortful.
130 and above
Often called the gifted range and used as a cutoff by high-IQ societies. Measurement precision is poor here because few items are hard enough to separate people cleanly.
Percentiles are the honest unit
A score of 110 is roughly the 75th percentile; 120 is about the 91st; 130 is near the 98th; 140 is around the 99.6th. Notice how quickly the percentile saturates at the top: the gap from 100 to 115 covers about a third of the population, while the gap from 130 to 145 covers under two percent of it.
This is why arguments about whether someone is a 135 or a 142 are close to meaningless. Both sit on the flattest part of the curve, where a handful of items decides the number.
The error bar you rarely see
Every score is an estimate with measurement error. Clinical batteries typically publish a 95% confidence interval of about five points in each direction. A short, unsupervised online test deserves a wider band — often ten points or more.
Read your result as a range with the reported number in the middle. Chasing a two- or three-point difference between two attempts is chasing noise, and no single score should drive a decision about study, work or self-worth.
Why the old labels were retired
Twentieth-century manuals attached clinical labels to low bands and words like "superior" to high ones. Several of those terms became insults, and all of them encourage reading a percentile as an identity. Current practice prefers a plain statement: "performed better than about 84% of the reference sample on non-verbal reasoning." Longer, and the only version that is actually true.
Ask which population you are compared to
A percentile is meaningless without a reference group. Clinical tests are normed on samples chosen to mirror a national population. Online tests, including ours, are normed on whoever visits — a younger, more curious, puzzle-friendly group. Same number, different meaning, and any site that will not tell you which one it used is not worth trusting.