Wonderlic to IQ: how the scores compare

Short answer: Wonderlic publishes no IQ conversion. None of the six Wonderlic pages we read reports an IQ, and its current hiring test, Wonderlic Select, gives candidates no single cognitive number at all. You can make a rough estimate by placing a raw score on the publisher's 2002 norm and reading off the same percentile on the IQ scale. By our calculation, a raw 20 lands near 97 and a raw 30 near 116. Treat any such figure as an estimate. By our calculation from McKelvie's 1992 reliability study, retest error on the Wonderlic alone gives a 90% range of about 9 points either way. The gap to a full IQ test can be larger: Dodrill's 1980 study found a 90% probability that a Wonderlic estimate fell within 10 points of the WAIS score, so about 1 in 10 missed by more.

Wonderlic to IQ: estimated equivalents

These estimates are our calculation. Wonderlic publishes no such chart. They use the publisher's 2002 norm for the adult working population, an average of 21.75 and a standard deviation of 7.60, as reported by Buehl and Melchers in a 2017 paper in Frontiers in Psychology. We turn each raw score into a z-score, then place it at the same point on the IQ scale, which has an average of 100 and a standard deviation of 15. Each line shows how many people in that norm group score lower, which is what a percentile means, worked out from the normal curve.

• Raw 10: higher than about 1 in 16 of the norm group, IQ-scale estimate about 77
• Raw 15: higher than about 1 in 5, estimate about 87
• Raw 20: higher than about 2 in 5, estimate about 97
• Raw 25: higher than about 2 in 3, estimate about 106
• Raw 30: higher than about 6 in 7, estimate about 116
• Raw 35: higher than about 24 in 25, estimate about 126
• Raw 40: fewer than 1 in 100 score higher, estimate about 136

Read each line as a range. By our calculation from a retest reliability of .872 (McKelvie, 1992), a 90% band for retest error on the Wonderlic alone runs roughly 9 IQ points either side of each estimate. The gap between an estimate and a full IQ test can be larger. Dodrill's 1980 study found a 90% probability that a Wonderlic estimate fell within 10 points of the WAIS Full Scale IQ, so about 1 in 10 missed by more. The list also assumes the Wonderlic norm group looks like the general adult population that IQ tests are normed on. It does not quite, as the sections below explain. Our guide to IQ percentiles shows how a percentile maps onto the IQ scale.

One score, four answers

Take a raw score of 30 and run it through each method you will find online.

1. Multiply by 5: 150. A test-prep site, JobTestPrep, gives this rule and cites no source for it.
2. Multiply by 2, add 60: 120. A 2011 SB Nation fan post, Blogging The Boys, gives this rule and cites no source either.
3. Percentile equating on the 2002 norm: 116, by our calculation above.
4. Prediction adjusted for the correlation: about 115 if the two tests correlate at .92, and about 109 if they correlate at .55, by our calculation.

The first two have no published source we could find. The last two are our own arithmetic from published figures, and the next sections show where each input comes from. The spread runs from 109 to 150 for one score.

What the Wonderlic is

The Wonderlic is a short timed test of general mental ability used mostly in hiring. Its history page says E.F. Wonderlic, "Director of Personnel for Household Finance in 1932," wanted a faster way to measure ability, and "he also created the first short-form intelligence test."

The cognitive part has stayed the same size. Wonderlic's sample-questions page (updated March 2026) says the full Cognitive Ability assessment "will contain 50 questions" and that "You will have exactly 12 minutes to provide as many correct answers as you can." No calculator or dictionary is allowed. Your raw score is the number of correct answers, so it runs from 0 to 50.

The product around it has changed. Wonderlic's candidate FAQ (updated March 2026) says Wonderlic Select, formerly Wonscore, has three sections, "one timed and two untimed," and "requires an average of 35 minutes to complete." The timed part is the 50-question cognitive section. The other two cover motivation and personality.

Does Wonderlic publish an IQ conversion?

Not on any page we could find. We read six Wonderlic pages on 25 October 2026 (the homepage, History, About, Select, the candidate FAQ and the sample-questions page). None contains the term IQ or any score-to-IQ table. That is an absence check on six pages. It does not prove that no Wonderlic document anywhere has one.

The current test makes the question harder still. The candidate FAQ says "today's Wonderlic Select test does not provide a singular numerical cognitive ability score." The three parts are combined and judged against the demands of the job, and "traditional numerical scores for the cognitive portion of the assessment are not available." The same FAQ says there is no universal good or average Select score.

So if you took Wonderlic Select recently, you may have no raw score to convert. The estimates on this page apply to the classic 50-item raw score.

One research conversion did exist. Dodrill's 1980 study in Epilepsia estimated WAIS Full Scale IQs "from Wonderlic IQs" in its samples. The study's abstract does not give that conversion and we could not read the full paper, so it is not used here.

Where the popular formulas come from

Both folk rules agree at a raw score of 20, where each gives 100. They split quickly after that. A raw 10 is 50 under the times-5 rule and 80 under the times-2-plus-60 rule.

The times-2-plus-60 rule has a hidden logic. By our calculation, it is exactly percentile equating if the Wonderlic average were 20 and its standard deviation 7.5. Divide 15 by 7.5 and you get 2. Take 2 times 20 from 100 and you get 60.

The times-5 rule implies a Wonderlic standard deviation of 3, by the same logic. No source supports that. The publisher's 2002 norm puts the standard deviation at 7.60 (Buehl and Melchers, 2017).

The average of 20 is itself a popular figure rather than a published norm. Wikipedia says "a score of 20 is intended to indicate average intelligence," and a Colts.com page says "a score around 20 is about average," but neither names a norm group. The only norm we found with a stated group and spread is the 2002 one, with an average of 21.75. By our calculation, the two assumptions differ by about 3 to 4 IQ points across most of the range. A raw 30 comes out at 116 on the 2002 norm and 120 on the 20-and-7.5 assumption.

The IQ bell curve explains why the standard deviation matters so much here.

How closely the Wonderlic tracks a full IQ test

Correlations between the Wonderlic and full-scale IQ tests run between about .55 and .93, depending on the study. A correlation of 1 would mean the two tests rank everyone in the same order.

• Dodrill, 1981: about .91 to .93 with the WAIS-R in 120 community subjects, as reported by Hicks, Harrison and Engle in the journal Intelligence in 2015. Hicks and colleagues say this single study is frequently cited as evidence of the Wonderlic's reliability and validity.
• Publisher's manual, 1992: .92 with the WAIS-R, as reported by Storek and Furnham in the Journal of Genetic Psychology in 2013.
• Buckley, 1957: .65 with the original WAIS, also as reported by Hicks and colleagues, who note that Edinger and colleagues (1985) did not find correlations as large as Dodrill's.
• Matthews and Lassiter, 2007: .55 with the Woodcock-Johnson-Revised score for Broad Cognitive Ability in 37 college students aged 18 to 54, published in Psychological Reports. The Wonderlic was not significantly correlated with fluid reasoning in that sample (r = .26).

One caution sits behind the high figures. Hicks and colleagues point out that comparisons with the WAIS can be inflated by item content the two tests share.

Group averages can match closely while single people do not. Dodrill's 1980 study of adults with epilepsy (100 people, then 50 more) found the average error in estimating WAIS Full Scale IQ was 0.1 to 0.3 of a point. For one person, though, "there is a 90% probability that the Wonderlic will render an IQ score within 10 points of the WAIS Full Scale IQ." Saltzman and colleagues (1998) found relatively high agreement in a head-injury group as a whole, but not good agreement "on an individual case basis." Our article on the WAIS IQ test describes the clinical test these studies compare against.

Why an estimate lands closer to 100 than the formulas say

When two measures correlate imperfectly, an extreme score on one predicts a less extreme score on the other. Yu and Chen, writing in Frontiers in Psychology in 2015, describe this as "a phenomenon called regression to the mean," a term they trace to Galton in 1886. Linden, in BMC Medical Research Methodology in 2013, notes that "the RTM effects diminish as the correlation between pre-test and post-test increases."

Here is how that plays out for a raw 30, by our calculation on the 2002 norm. Percentile equating gives 116, the IQ score at the same percentile. If you instead want the best single guess of what that person would score on a full IQ test, you multiply by the correlation first. At .92 you get about 115. At .55 you get about 109.

The two numbers answer different questions. Equating says where the score sits. Prediction says what a full test would most likely show, and it always sits closer to 100. The times-5 rule goes the other way and stretches the scale, which is why it gives 150.

Speed and the 12-minute limit

The Wonderlic is a speeded test. Wonderlic tells candidates that "Many people do not finish all of the questions," and its FAQ says test takers "are not expected to complete all these questions in the time allotted."

How few finish has a historical figure only. Wikipedia quotes Wonderlic's 1939 article as saying the test was made long enough that "only about two to five per cent of average groups complete the test in the twelve-minute time limit." That figure is for the original forms and reached us second-hand. We found no current percentage.

Speed matters for any conversion. A raw score mixes how much you know with how fast you work, and the 50-item ceiling caps the top of the scale. Estimates at the high end of the list above are the least certain for that reason. Our piece on what an IQ test measures covers what each part of a full IQ test samples.

How reliable a Wonderlic score is

A Wonderlic score is fairly consistent from one sitting to the next. McKelvie, in the Journal of General Psychology in 1992, gave the test twice over three weeks to 302 undergraduates, with the second sitting run under several different conditions. In the standard test-retest condition, the reliability was .872. Wikipedia says that in 1982 Dodrill gave the Wonderlic to 57 adults twice over a five-year period and found a reliability of .94.

Consistent is not exact. By our calculation, a reliability of .872 and the 2002 standard deviation of 7.60 give a standard error of about 2.7 raw points, or about 5.4 points once moved to the IQ scale. At .94 the figure is about 3.7 IQ points. In Hall v. Florida (2014), the Court's opinion says that for most IQ tests, "the SEM means that an individual's score is best understood as a range of scores on either side of the recorded score." That applies to every score on this page.

What an estimate can and cannot tell you

An estimate can tell you roughly where your raw score sat against the publisher's working-adult norm. It cannot tell you your IQ.

• It is not a clinical result. A psychologist giving a full test such as the WAIS measures more abilities over a longer session. A 12-minute screen converted by arithmetic is not the same thing.
• The norm group shifts. Sackett and Ostgaard, in the Journal of Applied Psychology in 1994, compared applicant pools for 80 jobs. For jobs above the lowest level of complexity, pool standard deviations "average 10% lower than the broad norm group SDs" in their data. So the same raw score can sit at a different percentile depending on who you are compared with.
• It is not on Mensa's list. American Mensa says it accepts scores from "approximately 200 different standardized intelligence tests," and its published list is not complete. The Wonderlic does not appear on that list. The page does not say Mensa rejects it.
• It is not a tool for assessing injury. Saltzman and colleagues (1998) concluded the test "is not a suitable tool for psychological assessment of individuals with known or suspected head injury."

If you need an IQ score for a school, a clinic or a legal matter, ask a psychologist about a full assessment. Our guide on whether IQ tests are accurate explains why two tests can give you different numbers.

Taking a test yourself

If you want a score that is reported on the IQ scale from the start, a matrix reasoning test is one option. The WhatIQ test has "24 picture puzzles, about 15 minutes," gives a score on the standard IQ scale with the range it most likely sits in, and calls the result "an estimate from 24 puzzles, not a clinical assessment" (whatiq.com, read 25 October 2026).

Frequently asked questions

Is the Wonderlic an IQ test?

Not in the clinical sense. It is a 12-minute test of 50 questions that Wonderlic's history page calls the first short-form intelligence test. The six Wonderlic pages we read on 25 October 2026 do not mention IQ, and the current Wonderlic Select gives candidates no single cognitive number. Its score correlates with IQ tests, but it does not report an IQ, and an estimate made from it is not a clinical result.

Does the Wonderlic correlate with IQ?

It does, though how strongly depends on the study. Dodrill (1981) found about .91 to .93 with the WAIS-R in 120 adults, as reported by Hicks and colleagues in 2015. Matthews and Lassiter (2007) found .55 with the Woodcock-Johnson-Revised in 37 college students. Dodrill's 1980 epilepsy study found a 90% probability that an individual estimate fell within 10 points of the WAIS score.

What is a good Wonderlic score?

Wonderlic's candidate FAQ (updated March 2026) says there is no universal good or average score on today's Wonderlic Select. For the classic raw score, the publisher's 2002 working-adult norm had an average of 21.75 and a standard deviation of 7.60, as Buehl and Melchers reported in 2017. Many pages call 20 average without naming a norm group. What counts as good depends on the job and the applicant pool.

How rare is a score of 40 on the Wonderlic?

Rare on the classic test. On the publisher's 2002 norm (average 21.75, standard deviation 7.60, as reported by Buehl and Melchers in 2017), a raw 40 leaves fewer than 1 in 100 working adults scoring higher by the normal curve, an IQ-scale estimate near 136 by our calculation. Wonderlic's 1939 article, as quoted by Wikipedia, said the test was made long enough that "only about two to five per cent of average groups complete the test in the twelve-minute time limit."

Is the Wonderlic still used?

Yes, mainly by employers. Wonderlic's candidate FAQ (updated March 2026) says organizations use Wonderlic Select as part of their hiring processes. The classic test still appears in research too: a 2026 PLoS One study gave participants 12 minutes to answer up to 50 Wonderlic items. The NFL stopped giving it at the Combine for 2022, CBS Sports reported in January 2022, and Sports Illustrated confirmed it was not given at the 2022 event.

Can you convert an IQ score to a Wonderlic score?

Only as a rough estimate. Running our calculation in reverse on the publisher's 2002 norm (average 21.75, standard deviation 7.60, as reported by Buehl and Melchers), an IQ-scale score of 116 lines up with a raw score of about 30. The same limits apply both ways: Wonderlic publishes no chart, the tests correlate imperfectly, and any score carries a range of several points.