IQ Bell Curve Calculator
Enter any IQ score to get its exact percentile, its rarity as 1 in N, how many standard deviations it sits from the mean, and its position on the bell curve.
Educational statistics tool. This plots where a score falls on a normal distribution. It is not an IQ test or a psychological assessment, and it can't measure anyone's intelligence. For a real evaluation, consult a licensed professional.
Where does your IQ fall on the bell curve?
IQ scores follow a normal distribution with mean 100 and standard deviation 15 (Wechsler scale). Here's where common scores land:
- IQ 70: 2.3rd pctl (1 in 44)
- •
- IQ 85: 16th pctl (1 in 6)
- •
- IQ 100: 50th pctl (1 in 2)
- •
- IQ 115: 84th pctl (1 in 6)
- •
- IQ 130: 97.7th pctl (1 in 44)
- •
- IQ 145: 99.9th pctl (1 in 741)
Enter any score in the calculator to see its exact percentile and rarity, and to visualize it on the bell curve. Works with any mean and standard deviation.
IQ Score Percentile Reference Chart
Standard deviations, percentiles, and rarity for common IQ values (mean 100, SD 15):
| IQ Score | Z-Score | Percentile | Rarity | Classification |
|---|---|---|---|---|
| 55 | −3.00 | 0.1st | 1 in 741 | Extremely Low |
| 70 | −2.00 | 2.3rd | 1 in 44 | Borderline |
| 85 | −1.00 | 16th | 1 in 6 | Low Average |
| 100 | 0.00 | 50th | 1 in 2 | Average |
| 115 | +1.00 | 84th | 1 in 6 | High Average |
| 120 | +1.33 | 91st | 1 in 11 | Superior |
| 130 | +2.00 | 97.7th | 1 in 44 | Gifted |
| 145 | +3.00 | 99.9th | 1 in 741 | Genius |
| 160 | +4.00 | 99.997th | 1 in 31,574 | Profoundly Gifted |
Jump to: Percentile · Rarity (1 in N) · Standard deviations · Scale converter · Quick answers
IQ to Percentile
A percentile is the share of the modeled population scoring at or below you. On the Wechsler scale (mean 100, standard deviation 15), an IQ of 100 is the 50th percentile by definition. 115 is the 84th. 130 is the 97.7th. The calculator gets there in one step: it computes z = (IQ minus 100) / 15, then reads the percentile off the normal cumulative distribution function, the same CDF every stats textbook and scipy.stats use.
Common scores on SD 15:
- IQ 85 is the 16th percentile.
- IQ 100 is the 50th.
- IQ 115 is the 84th.
- IQ 120 is the 91st.
- IQ 130 is the 97.7th.
- IQ 145 is the 99.9th.
Going the other way, from a percentile back to a score, switch the tool to "IQ from Percentile" and it inverts the CDF for you. Ask for the 90th percentile and you get roughly 119. Ask for the 99th and you get about 135.
How Rare Is an IQ Score? (1 in N)
Rarity is the tail probability turned into odds. Take the fraction scoring at or above a value, then divide it into 1. IQ 130 sits at the 97.7th percentile, so about 2.3% score higher. One divided by 0.0228 is close to 44, which is why an IQ of 130 comes out around 1 in 44. Climb higher and the tail thins fast.
- IQ 115: about 1 in 6 score at or above.
- IQ 120: about 1 in 11.
- IQ 130: about 1 in 44.
- IQ 140: about 1 in 261.
- IQ 145: about 1 in 741.
- IQ 160: about 1 in 31,574.
Those tail figures are also where the model quietly stops describing anything real. Past roughly IQ 145 you're slicing the population thin enough that no standard test carries the items to resolve one person from the next, so read a 1-in-30,000 number as what the curve predicts, not what a testing session can measure.
Standard Deviations and z-Scores
A z-score answers one question. How many standard deviations from 100 is this score? On SD 15 it's a quick mental calc: 115 is z +1, 130 is z +2, 145 is z +3. Below the mean the sign flips, so 85 is z minus 1 and 70 is z minus 2.
That's what makes the 68-95-99.7 rule click. Roughly 68% of scores land within one SD of the mean, meaning 85 to 115. About 95% fall within two SDs, 70 to 130. And 99.7% sit within three, 55 to 145. If you want the exact figures the calculator returns, they're 68.27%, 95.45%, and 99.73%.
Converting Between Test Scales
Not every test uses SD 15. The Wechsler tests (WAIS, WISC) do. The older Stanford-Binet Form L-M used SD 16. Cattell used SD 24. Because the spread differs, one raw number means different things across tests, and one percentile lands on different raw numbers.
The clean fix is to convert through the z-score instead of the raw score. Compute z on the source scale, z = (IQ1 minus 100) / SD1, then rebuild it on the target: IQ2 = 100 + z times SD2. Run the 98th percentile (z is about 2.05) through it and you get roughly IQ 131 on Wechsler, 133 on Stanford-Binet, and 149 on Cattell. Same person, same standing, three different labels.
This is also why "Mensa requires 130" needs a footnote. Mensa admits the top 2%, and it lists a qualifying score of 130 on the Wechsler scale or 132 on the Stanford-Binet. Both are the same 98th-percentile cutoff wearing a different SD. Switch the tool to Scale Converter to move any score between the three.
Quick Answers
What percentile is an IQ of 130? IQ 130 is the 97.7th percentile on the SD-15 scale. About 2.3% of people score at or above it, roughly 1 in 44.
What percentile is an IQ of 120? IQ 120 is about the 91st percentile, a z-score of 1.33. Around 9% score higher, close to 1 in 11.
What percentile is an IQ of 145? IQ 145 is the 99.9th percentile, a z-score of 3. Roughly 1 in 741 people reach it.
How rare is an IQ of 140? IQ 140 is the 99.6th percentile, about 1 in 261.
What does an IQ of 100 mean? 100 is the built-in center of the scale, not a count of correct answers. It's the 50th percentile by design, so half the reference group scores below it and half above.
What is the 68-95-99.7 rule for IQ? About 68% of scores fall between 85 and 115, 95% between 70 and 130, and 99.7% between 55 and 145. Those bands are one, two, and three standard deviations from 100.
What the Bell Curve Actually Shows You
Someone tells you they scored 120 on an IQ test and you nod along, but what does 120 actually buy you? On its own, a raw score is just a number. Drop it onto a normal distribution and it turns into something you can read. A 120 lands about 1.33 standard deviations above the mean of 100, which puts it near the 91st percentile. An IQ bell curve visualizer does that drop for you, showing at a glance how common or rare a score is against the rest of the modeled population.
One thing to be clear about up front: this isn't a test. No questions, no timer, no result waiting at the end. You type in a number, any number, and the tool plots it on a Gaussian curve so you can watch how z-scores, percentile ranks, and areas under the curve actually behave. IQ is just the familiar example here, the way a physics class reaches for a frictionless ramp. Convenient enough to make the math stick, simple enough that the concept stays in focus. The subject is statistics, not the people behind the numbers.
How to Use It
- Enter one or more scores. Type any number: 85, 100, 130, whatever you want to explore.
- Check the mean and SD. Defaults are 100 and 15 (the Wechsler scale). Change them if you want to model a different test or distribution.
- Read the output. The curve appears with your score marked, plus a z-score, percentile rank, and shaded area showing what fraction of the population falls below that point.
You can plot multiple scores at once to see how they sit relative to each other. You can also highlight standard-deviation bands. The region from 85 to 115 (one SD on either side of the mean) captures roughly 68% of the distribution. Widening to two SDs (70 to 130) covers about 95%. These are the numbers behind the famous 68-95-99.7 rule, and seeing them shaded on a curve makes the abstraction concrete.
Under the Hood
Every output boils down to one formula: z = (x − μ) / σ. That is the z-score, how many standard deviations your value sits from the mean. Once you have the z-score, the tool feeds it into the cumulative distribution function (CDF) of the normal distribution to get the percentile. A z of 0 gives the 50th percentile. A z of +1 gives about the 84th percentile. A z of +2 gives about the 97.7th.
The curve itself is drawn from the probability density function: f(x) = (1 / (σ√(2π))) × e^(−(x−μ)²/(2σ²)). You never compute this by hand. The point is to see its shape. The peak sits at the mean, the curve falls off symmetrically on both sides, and the tails stretch out but never quite touch zero. Changing σ squeezes or spreads the curve; changing μ shifts it left or right. Play with both and you will feel how distributions behave far better than memorizing formulas ever teaches.
Worked Example
You are studying for a stats midterm and want to understand where a score of 112 falls on a standard IQ bell curve (mean = 100, SD = 15).
Input: Score = 112, Mean = 100, SD = 15
Z-score: (112 − 100) / 15 = 0.80
Percentile: CDF(0.80) ≈ 78.8th percentile
Interpretation: A score of 112 is 0.8 standard deviations above the mean, higher than roughly 79% of the modeled population.
Now compare that to a score of 95: z = (95 − 100) / 15 = −0.33, which lands near the 37th percentile. The five-point jump from 95 to 100 moves you about 13 percentile points, while the twelve-point jump from 100 to 112 moves you about 29 percentile points. That uneven relationship between raw-score gaps and percentile gaps is one of the most important things the bell curve teaches. Small differences near the center of the distribution shift percentiles less dramatically than the same differences further out.
Watch Out For
- Different scales use different SDs. Wechsler tests use SD = 15. The Stanford-Binet historically used SD = 16. Older tests sometimes used SD = 24. Take a raw score of 130. On an SD-15 scale that is z = 2.0 and the 97.7th percentile, about 1 in 44. On an SD-16 scale the same 130 is only z = 1.875, closer to the 97th percentile and about 1 in 33. Comparing scores across tests without matching the SD is how people end up off by a full rarity bracket.
- Percentile does not mean percent correct. The 90th percentile does not mean you got 90% of answers right. It means you scored higher than about 90% of the reference population. These are completely different concepts, and confusing them is one of the most common mistakes in introductory statistics.
- Ceiling effects at extremes. The normal distribution is a mathematical model. Real tests have a finite number of questions, so they cannot distinguish between very high or very low scores with the same precision as mid-range scores. A score above 145 or below 55 pushes into territory where the model is more theoretical than practical.
- Cultural and contextual bias. IQ tests are standardized against specific reference populations. Scores reflect performance on those particular tasks under those particular conditions, not some universal or fixed measure of ability. This visualizer shows the math of distributions, not the psychology of intelligence. For anything beyond learning statistics, consult a qualified professional.
Straight Answers
Is this an IQ test? No. There is no assessment here. You enter numbers and the tool plots them on a distribution curve. It is a statistics learning aid, not a psychological evaluation.
Why does a 5-point difference near 100 matter less than a 5-point difference near 130? Because the bell curve is steepest near the mean and flattens toward the tails. Near the center, lots of people cluster within a few points, so moving 5 points does not change your percentile much. Near the tails, the population thins out quickly, so the same 5-point shift crosses a bigger percentage of the remaining distribution.
Can I use different mean and SD values? Yes. Change the mean to 500 and SD to 100 and you have roughly the SAT distribution. The math is identical. Only the labels on the axis change.
What is the 68-95-99.7 rule? In any normal distribution, about 68% of values fall within 1 SD of the mean, 95% within 2 SDs, and 99.7% within 3 SDs. On a standard IQ curve, that translates to 85 to 115, 70 to 130, and 55 to 145.
Why does the tool warn about extreme scores? Above about z = 3 (IQ 145 on SD 15) the normal model is doing more extrapolating than describing. Real tests run out of items to separate people that far into the tail, so the percentile stays mathematically clean but gets practically fuzzy.
Educational tool. Results are estimates.
Educational only. These comparisons use public data and general models. Verify anything decision-critical against current local sources.
Frequently Asked Questions About IQ Bell Curves
What does the IQ Bell Curve Visualizer show?
It displays a normal distribution (bell curve) using IQ scores as an example, showing how scores distribute around a mean of 100 with a standard deviation of 15. You can see z-scores, percentiles, and shaded regions representing population proportions. This is a math learning tool, not a psychological assessment.
Why is the mean 100 and standard deviation 15?
These are the conventional parameters for standardized IQ test scoring systems. Mean 100 represents the center or 'average' in the model, and SD 15 means about 68% of scores fall between 85 and 115. Different standardized tests may use different means and SDs.
How are percentiles calculated from IQ scores?
Percentiles are calculated using the cumulative distribution function (CDF) of the normal distribution. For example, a score of 100 (z = 0) is the 50th percentile, meaning 50% score below it. A score of 115 (z = 1) is approximately the 84th percentile.
Are percentiles exact for all standardized tests?
No. This tool uses the idealized normal distribution model with specific mean and SD. Real test data may have slight variations. Always refer to official test documentation for precise percentile ranks and interpretations.
Can I use this to find my 'true' IQ?
No. This is an educational math tool for learning about normal distributions, z-scores, and percentiles. It is NOT a diagnostic psychological assessment. Real IQ testing requires professional administration, context, and interpretation.
What's the difference between a raw score and a standardized score?
A raw score is your original test performance (e.g., 45 out of 60 questions correct). A standardized score (like an IQ score) converts that raw score into a z-score or scaled score on a normal distribution, allowing comparison across different tests and populations.
Why is the bell curve symmetric?
The normal distribution model is mathematically symmetric, meaning equal areas above and below the mean. This is a theoretical model; real test score distributions may have slight asymmetries, but many approximate this shape well enough for teaching purposes.
How does changing the standard deviation affect the curve's shape?
A larger SD spreads the curve wider (more variability), making extreme scores less rare. A smaller SD makes the curve narrower and taller (less variability), with scores clustering tightly around the mean. The visualizer lets you experiment with different SDs.
Can I use this for other standardized tests, like SAT or GRE?
Yes, conceptually! Many standardized tests use normal models with different means and SDs. For example, SAT sections are often modeled with mean 500 and SD 100. You can adjust the visualizer's mean and SD to approximate those systems for learning purposes.
How can teachers use this tool for statistics lessons?
Teachers can demonstrate the 68-95-99.7 rule, show how z-scores translate to percentiles, compare multiple scores, and use shaded regions to teach probability and area under the curve. It makes abstract concepts concrete and interactive.
What is the 68-95-99.7 rule (Empirical Rule)?
In a normal distribution, approximately 68% of values fall within 1 SD of the mean, 95% within 2 SDs, and 99.7% within 3 SDs. For IQ: 68% score between 85 and 115, 95% between 70 and 130, and 99.7% between 55 and 145.
How rare are extreme scores (e.g., above 145 or below 55)?
Scores beyond ±3 SDs from the mean are very rare. For example, an IQ above 145 (z > 3) occurs in less than 0.3% of the population under the normal model. The bell curve's tails drop off rapidly, making such scores increasingly uncommon.
Is this tool a real IQ test?
Absolutely not. This is a math and statistics learning tool that uses IQ scoring conventions as a familiar example. It cannot measure intelligence, cognitive ability, or any psychological trait. For real assessments, consult a licensed professional.
What's the difference between the model and reality?
The model is an idealized, smooth bell curve. Real test score data may have slight bumps, skewness, or measurement error. The model is useful for teaching core concepts but should not be treated as a perfect description of all real-world test distributions.
How can I avoid misusing this tool?
Use it for learning statistics: z-scores, percentiles, normal distributions. Do NOT use it to label people, make judgments about intelligence, or replace professional psychological evaluation. Always emphasize it's a teaching simplification, not a diagnostic instrument.
Was this calculator helpful?
Your rating helps us improve every EverydayBudd tool.
Explore More Statistics & Probability Tools
Explore More Fun & Interactive Tools
Discover our collection of interactive visualizers, simulators, and educational tools for learning and entertainment