IQ by Profession: The Average Scores Behind the Jobs, and Their Limits

How this list was built

Every figure on this page comes from a published study or a test publisher's data, attributed by name. We excluded the unsourced tables that circulate online, which typically assign a single IQ to dozens of jobs with no sample, no test and no date.

Three bodies of evidence survive that filter. The first is the Army General Classification Test, given to American recruits in World War II and analyzed by occupation in the 1940s. The second is the Wonderlic Personnel Test, an employment screening test, whose 1992 applicant data the psychologist Linda Gottfredson converted to IQ equivalents in 1997. The third is the Wisconsin Longitudinal Study, which followed the high school class of 1957 for decades.

All three report group averages. None describes what any individual in a job can do.

The Army data: World War II recruits

The most-cited occupational table comes from the war. Naomi Stewart, writing in the journal Occupations in 1947, grouped the AGCT scores of more than 250,000 Army inductees by 227 civilian occupations. A 1952 University of Detroit thesis by Trygg Engen reproduced her ten highest-ranking occupations and their medians.

The AGCT used its own scale, with a mean of 100 and a standard deviation of 20, as Dael Wolfle and Toby Oxtoby described in 1952. The right-hand column rescales each median to the modern 15-point standard deviation, which assumes the Army population stands in for the general one. It does not fully: the sample was wartime Army inductees, screened for service and tested in the 1940s. Half of the top ten are students, not working professionals.

A parallel study by Thomas W. Harrell and Margaret S. Harrell, "Army General Classification Test Scores for Civilian Occupations," appeared in Educational and Psychological Measurement in 1945. We could not access its full table and report none of its figures here.

The Wonderlic data: job applicants in 1992

Linda Gottfredson's 1997 paper "Why g Matters," published in Intelligence, plotted Wonderlic scores by the job applied for, drawing on the publisher's 1992 manual. She converted the median applicant's score to an IQ equivalent and treated it as the level needed to be competitive.

These are applicants, not employees, and they are people who chose to apply. For the test's most public use, see our guide to the highest Wonderlic scores in the NFL.

Degrees and fields: the 1951 college seniors

A different window comes from education. Wolfle and Oxtoby, working for the Commission on Human Resources and Advanced Training, reported in 1952 on 38,420 college seniors who took the Selective Service College Qualification Test in 1951, a self-selected group of men, with scores converted to the AGCT scale.

The same report put the average bachelor's graduate near 126 on the AGCT scale, graduate students near 129 and science PhDs near 138, about 129 on today's scale. PhD medians in chemistry, earth sciences, engineering, physics and psychology ran from roughly 138 to 143. The gap between the top and bottom fields is about eight AGCT points, less than half of one standard deviation.

The spread inside every job

The averages hide the main finding. "There is much IQ variation within all occupations and much overlap among them," Gottfredson wrote. In her data, the middle 50 percent of applicants to a single job generally covered 15 to 20 IQ points.

Robert Hauser of the University of Wisconsin reached the same place from a different angle. His 2002 paper drew on the Wisconsin Longitudinal Study: a one-third random sample of the state's high school class of 1957, 4,994 men and 5,323 women, tested with the Henmon-Nelson Test of Mental Ability in their junior year. He charted ability distributions for 65 occupation groups among men and 62 among women. The correlation between occupational standing and test score was "no more than moderate: 0.39 to 0.44." The charts did show an occupational hierarchy, he noted, but with heavy overlap from top to bottom.

Put plainly, a bright custodian and an average accountant are common. Knowing someone's job tells you little about that person's score.

What IQ adds once you are hired

Test scores help sort people into occupations more than they decide success within one. Tarmo Strenze's 2007 meta-analysis found a corrected correlation of 0.43 between intelligence and occupational status, close to the 0.38 for parents' socioeconomic status.

For performance on the job, the estimate has come down. Paul Sackett and colleagues, in the Journal of Applied Psychology in 2022, put the validity of cognitive ability tests at 0.31, a useful predictor but behind structured interviews, job knowledge tests, biographical data and work samples. Experience, skill, conscientiousness and the job itself carry much of the load.

How to read job averages honestly

Treat every number here as a historical, group-level snapshot. The Army data are 80 years old and cover wartime inductees only. The Wonderlic figures describe applicants in 1992. Raw scores have since risen and tests have been renormed, so none of these medians maps exactly onto a test taken today.

The pattern is still useful. Jobs heavy in abstract reasoning and paperwork drew higher medians; jobs built on routine physical tasks drew lower ones. To see where scores fall on the modern scale, compare IQ 120 with what counts as a good IQ score and the top 1 percent. For a rough personal reading, our free IQ test is a practice screener; why online tests fall short explains its limits. More in our IQ guides.

Editorial standard: IQ figures for historical figures are retrospective estimates and are labeled as such; scores for living people appear only when the person disclosed them or a reputable outlet reported them. Corrections: contact the editors.