What does a percentile rank mean?
After receiving a score, people almost always ask, “Is my score good?” GMAC does not answer with a cutoff. It reports a percentile rank, which it defines as the share of test takers you outperformed. A rank of 75% means you scored higher than 75% of test takers.
A percentile is not a count of correct answers or a “score out of 100.” It places a total within a population over a given period. The same total of 655 may be interpreted differently if the reference population changes, which is why the table specifies a data period, shown below.
The GMAT total ranges from 205 to 805 in 10-point increments and always ends in 5. Every possible total therefore has a rank in the official table, which is what shows where you stand—not a figure you heard elsewhere.
How do you read a few rows of the table?
GMAC’s concordance table ranks totals on the current exam by percentile. The table below reproduces six rows, rounded to one decimal place as in the source. It does not replace the full table, which lists every increment.
Two comparisons are useful. From 555 to 565, a 10-point increase in the total moves the rank up 3 percentage points, from 47.4% to 50.4%. From 565 to 575, the same increase moves it up 6.3 points, from 50.4% to 56.7%. Higher up, the scale compresses: from 705 to 755—a 50-point increase in the total—the rank rises 2.3 points (from 97.5% to 99.8%), and several consecutive totals round to 100.0%.
| Total | Percentile rank | Interpretation |
|---|---|---|
| 205 | 0.2% | Bottom of the scale |
| 555 | 47.4% | Outperformed just under half of test takers |
| 605 | 69.4% | Outperformed about seven in ten test takers |
| 655 | 89.6% | Outperformed about nine in ten test takers |
| 705 | 97.5% | Outperformed more than 97 out of 100 test takers |
| 805 | 100.0% | Top of the scale, shared with several neighboring totals |
Why should you interpret a 10-point gap with caution?
Between 625 (78.2%) and 635 (80.9%), the table shows a rank difference of less than three points. More importantly, GMAC urges caution when directly comparing two test takers’ scores: there is a standard error for the difference between two totals. On its admissions page, GMAC puts it at 41 points, above the 29-point standard error of measurement, and notes that scores 40 to 50 points apart may reflect the same performance in a program.
In practice, the gap between a 635 and a 625 alone is not enough to establish a clear difference in ability. Interpret the result in light of its uncertainty and the rest of the application. This caution does not exempt you from meeting a minimum explicitly set by a program: measurement error does not change its admissions rule.
The score calculation, in the GMAT guide GuideA table that changes each year
GMAC says the table values may be updated each year in the third quarter. The version reviewed on October 15, 2026, was published in August 2026; its ranks are based on the period from July 1, 2021, to June 30, 2026. The PDF URL still contains “Jul2025,” but the publication date and data period appear at the foot of the document and identify the version.
A rank can therefore shift from one year to the next even if your total stays the same. Before citing a percentile, note the date of the table you consulted.
How do you place a score from the previous version?
The previous GMAT version, scored from 200 to 800, was retired on January 31, 2024. Since scores remain valid for five years, some of its totals still appear in applications. The two scales do not align directly: in the same table, GMAC publishes a concordance between totals on the two versions based on their percentile ranks.
Read it in two steps. Find the earlier total in the old-version column, then note every current total linked to it in the table. This is a percentile-based concordance, not an exact equivalence between two scores. An earlier total may map to several current totals; GMAC attributes this to the size of the score increments and observed frequencies. Keep the range, not its most favorable endpoint.
The clearest example is 700. On the old scale, it corresponds to current totals of 645 and 655, or a rank of 85.6% to 89.6%, while a current 705 ranks at 97.5%. Treating an old 700 as a 705 therefore overstates the rank by eight to twelve points. The percentile printed on an old score report depended on the table used at the time. To compare the two scales today, use the current table, and apply the same approach to a school requirement stated on the old scale.
| Previous version (200 to 800) | Current version (205 to 805) | Percentile rank |
|---|---|---|
| 750 | 695 to 715 | 96.9% to 98.6% |
| 700 | 645 to 655 | 85.6% to 89.6% |
| 650 | 595 to 615 | 66.3% to 75.4% |
| 600 | 555 to 565 | 47.4% to 50.4% |
| 550 | 515 to 525 | 31.6% to 34.0% |
| 500 | 485 to 495 | 21.7% to 25.4% |
The pitfall
Looking for a total that applies to every school
A statement like “below this total, there’s no point applying” creates a false sense of certainty. No percentile rank tells you what a specific program requires: one program may target a high rank, while another publishes no requirement. A percentile describes your position among test takers; it does not predict admission.
The opposite pitfall is to rely on an impressive rank without reading the admissions page. A figure without a source, repeated on a forum or by a test-prep company, does not replace the rule published by the program itself.
How do you read percentiles by section?
Each of the three sections—Quantitative Reasoning, Verbal Reasoning, and Data Insights—is scored from 60 to 90, and each contributes equally to the total. GMAC also publishes section-level concordance tables for Quantitative Reasoning and Verbal Reasoning. For Data Insights, ranks are based on the previous three years of test administrations.
Read each section using its own rank, not the total’s. Two test takers with the same total may have different profiles, and a program may weigh one section more heavily than the others. A solid total can hide one section that is much lower: the differences across the three lines tell you more. The Score Report article offers a review framework to turn that information into focused practice.
How do you assess your score for a specific program?
Start with the program, not the number. Open its admissions page and class profile, record the published benchmarks along with the link, application cycle, and date reviewed, then place them in GMAC’s table. Distinguish a required minimum from a recommended score and a class average or median: the latter are not eligibility requirements. If the pages you consult give no benchmark, ask the admissions office.
Then compare those benchmarks with your total while accounting for measurement uncertainty. Do not infer your probability of admission or a guaranteed score increase from another attempt. Meet any explicitly required minimum; for other benchmarks, also consider your section scores, the rest of your application, and the application calendar before deciding whether to retake the test or send your report.
No school cutoff is given here because it varies by program, admissions route, and application cycle. This article also does not promise that a retake will raise your score.
Key takeaways
- A rank of 75% means you scored higher than 75% of test takers during the table’s data period.
- A 10-point gap alone is not enough to establish a clear difference in ability between two test takers.
- The table may be updated each year: cite its version date with the rank, and place an earlier score using the current concordance table rather than comparing totals directly.
- Assess your score against your target program: a minimum, a recommendation, and a class average mean different things.