How Much Higher Can GMATs Possibly Go?

ranking careers

It’s common for every generation to think that if they had to compete with the current crop of applicants, they would never make it into their alma maters. For many, however, it’s no joke.

Consider that only 20 years ago, not a single entering MBA class at any business school boasted a class average GMAT of 700 or more. In fact, the average GMAT score for a Top Ten U.S. program was just 669, with a high of 690 at Stanford and a low of 650 at MIT Sloan (see table).

For the group of students who entered these same ten schools this fall, the average GMAT score bloated to a 725 average, with a high of 733 at Stanford and a low of 715 at both Columbia Business School and UC-Berkeley’s Haas School of Business. That’s a 56-point increase in average GMATs for the Top Ten schools in the past 20 years.


The school’s with the biggest increases? Wharton leads the group with average GMAT scores that are now 70 points higher than they were in 1996: 732 for this year’s entering class versus 662 two decades ago. MIT Sloan’s average GMAT ballooned to 716 from 650, up 66 points. Northwestern University’s Kellogg School of Management saw its average class GMAT go to 724 from 650, a 64-point jump.

What’s behind the jump and how much higher can these scores get? A number of factors. For one thing, average scores on the GMAT test over the past two decades have gone up considerably. In testing year 1995, for example, the average score was 503. In testing year 2015, it had ballooned to 554, an increase of 51 points (see table below).

For another, some highly admired schools have clearly adopted admission policies that put greater emphasis on high GMAT scores. Admission officials are often under pressure to go for higher scores because U.S. News & World Report includes average GMATs as part of its methodology to rank the best MBA programs. They also believe that higher reported scores enhance the notion that a school is attracting the best possible candidates, even though there is little to no correlation that those who score highly on the GMAT will have successful careers.


“There’s definitely a ‘chicken and egg’ aspect to all of this,” explains Dan Bauer, founder and CEO of The MBA Exchange, a leading MBA admissions consultant. “Schools with higher median GMAT scores are perceived as having higher quality MBA programs. So, the battle to improve and defend school rankings is only escalating. The most controllable and visible way to elevate ranking is by increasing the median GMAT score for admits. This, in turn, attracts applicants with higher GMAT scores, thus raising the median scores at the schools that admit them.”

Studying for the test itself has also become a more effective exercise. “The availability and affordability of test preparation resources are expanding by the minute,” adds Bauer. “A Google search on ‘GMAT tutoring’ produces over a quarter-million listings. Do-it-yourself software, local and online classes, one-on-one personalized guidance, and mass-market books are flooding the market. In addition to having expert tutors, our firm even features a licensed psychologist who specializes in test-anxiety counseling. Today, MBA applicants worldwide have an unprecedented array of resources to help them study, practice and improve their GMAT performance.”

Source: Graduate Management Admission Council

Source: Graduate Management Admission Council

  • A_Sound_Bite

    The test IS a little easier, yes?

  • avivalasvegas

    I believe that some of the skills you mention are not correlated to the GMAT score at all, which is why the higher score doesn’t not necessarily reflect a better candidate.

    There are many abilities that a higher GMAT score DOES reflect rather accurately… but I do not believe that these qualities alone comprise a better MBA candidate.

    True quality comes when you admit people with varied, deep and diverse experiences with the highest scores possible.

  • MistaFabro

    Unless you’re assuming that things like business skills, social skills, etc. are negatively correlated to GMAT score (a questionable assumption), I see no reason not to think that, on average, you wouldn’t get more quality from a 98th percentile scorer than from a 90th percentile scorer.

    That’s obviously not to say that every 750 scorer is better than every 700 scorer, but on average over a large pool I think it’s fair to say you’re better off admitting more 750’s.

  • Person

    Not quite. A 690 used to be roughly the 90th %ile. Now, a 710 is 90th %ile. A 730 is 96th %ile. So schools have actually been taking people with higher scores that they previously would have denied over other candidates with lower scores but other factors. Some element of the GMAT rise (maybe 30-40%) is attributable to a percentile shift. The rest is due to a shift in strategy.

  • avivalasvegas

    I would be interested to see evidence that a 750 GMAT admit brings more quality to the table than a 700 GMAT student as I would be inclined to think otherwise.

    Sure, its great to boost ranking metrics – that’s clear. But do Ad Coms have some sort of internal cut off, whereby any score above X is as good as another score above the same X.

  • JohnAByrne

    Unfortunately not but given the rise in the average score you’re probably close on your estimate.

  • Al


  • Al


    Do you have data on the percentages. Was a 690 today’s 730?

  • Travis Mahoney

    Great post, but the math on the last table is wrong… Harvard 725-650 = +75 increase, not 45

  • Barbara Coward

    Interesting to note that women test takers increased by 43 percent since 1995 while men test takers increased by 25 percent for the same period. Also, the mean total score for women increased by 13 percent since 1995 while the mean total score for men increased by 9 percent for the same period Evidence that outreach efforts by b-schools to attract more female candidates and help them succeed are working!

  • C. Taylor

    Could go higher than ~745-750 mean, but would mean brutal overemphasis on test prep and almost certainly sacrifice outstanding talent. Unreasonable for most of the class to extensively prep for the exam . . . Has anyone done decently accurate surveys of GMAT prep time for accepted candidates at the top?

    735 is as close to the limit as minimally reasonable and would already sacrifice fantastic talent, at present.