An IQ of 136 sits in the very high range, around the 99.18th percentile. Here is what that means, and what it does not.
Reference score
136
Percentile
99.18%
At or above
1 in 122
Reference scale
100 mean / 15 SD
Prepared by IQInsight · Calculation guide · Updated
A smaller upper tail does not mean a proportional increase in ability.
An IQ of 136 is 2.4 standard deviations above the reference mean. Its model percentile is about 99.18, and the upper tail is about 0.82%, roughly one in 122. At 135 the upper tail is about 0.98%, roughly one in 102. The one-point step looks small on the score scale but more noticeable when expressed as a reciprocal probability.
This is a useful example of why “one in” numbers need careful reading. Moving from roughly 102 to 122 changes the reciprocal by about a fifth, but that does not mean a fifth more intelligence. The comparison concerns the relative sizes of two small tails, not the amount of an ability in two people.
How the percentile is calculated.
The reference uses a mean of 100 and a standard deviation of 15. First convert the score to a z-score, then find the area below it on the standard normal curve.
Theoretical normal distribution. Mean 100, standard deviation 15. This does not measure your IQ or compare you with IQInsight users.
Put the number in perspective.
In a normal distribution with mean 100 and standard deviation 15, about 0.82% of scores fall at or above 136. That is approximately one in 122. The figure describes the reference model; it is not a count of people we have tested.
The quality of the test matters.
A number alone does not show how accurately it was measured. An established, professionally administered test has supporting evidence and defined interpretation rules. An online exercise without representative norms cannot establish the same conclusions.
Nearby scores on the same scale.
Theoretical percentiles, mean 100 and standard deviation 15
Use the comparison with 135 to understand probability, not to claim a percentage advantage in ability. A one-point difference has no established practical meaning without the original instrument’s precision evidence. This guide supplies the model calculation, not that evidence.
A score alone cannot tell you how a person thinks through a problem. This published sample illustrates a reasoning method; it is not calibrated to a 136 IQ level and does not predict a score.
Every club member plays chess. Some chess players compete nationally. What can be concluded about whether any club members compete nationally?
A. All club members must compete nationally
B. At least one club member must compete nationally
C. No club members can compete nationally
D. It cannot be determined
Reveal the answer and reasoning
Answer: It cannot be determined. National competitors could all be outside the club, or could include club members. Both arrangements satisfy the premises.
Check an alternative: All club members must compete nationally. The national competitors need not include every member of the club.
Try this approach. Try to construct a counterexample before deciding that an overlap is guaranteed.
This is one of our public sample questions. Seeing the solution may make a future attempt more familiar. Explore the sample report to see how answers are explained.
A few things you might be wondering.
Clear answers, before you begin.
Is 136 about 20% better than 135 because of the rarity figures?+
No. Reciprocal probabilities measure how often a threshold would be reached under a model. They are not units of ability. The actual score difference is one point, and interpreting it requires evidence about the test that produced the results.
Is 136 a good IQ score?+
On a theoretical scale with mean 100 and standard deviation 15, 136 is in the very high range, around the 99.18th percentile. Whether a particular test result supports that interpretation depends on the quality and standardization of that test.
How rare is an IQ of 136?+
Under that normal-distribution model, about 0.82% of values are at or above 136, approximately 1 in 122. This is a mathematical reference, not observed IQInsight user data.
Does this number describe all of my abilities?+
No. IQ scores do not describe every skill, interest, quality or circumstance. A score on an unstandardized online exercise is especially limited and should not be treated as a diagnosis or a prediction of personal success.
Can I compare scores from different tests?+
Only with care. Tests can differ in norms, scales, questions, administration and reliability. Similar-looking numbers are not necessarily equivalent measurements.
Can I get my result without paying?+
Yes. IQInsight is designed to provide the estimated score, theoretical percentile and all four domain breakdowns for free. The optional report explains the answers in more detail.
