GRE Scores and Percentiles: Reading Your Report

A GRE score report shows two numbers for each measure: a scaled score and, beside it, a percentile rank. The scaled score is the performance measurement — Verbal Reasoning and Quantitative Reasoning run from 130 to 170 in one-point steps, Analytical Writing from 0 to 6 in half-point steps. The percentile rank is that score's position in a crowd: ETS defines it as the percentage of test takers who scored lower than you. The ranks on current reports come from Tables 1B and 1C of the ETS GRE Guide to the Use of Scores, 2026–27 edition, and the crowd in question is everyone in the applicable reference group who took the test between July 1, 2022 and June 30, 2025.

Here is the single fact that makes GRE score percentiles worth understanding rather than skimming: identical scaled scores on different measures represent completely different standings. In the 2026–27 tables, a 160 on Verbal is the 82nd percentile while a 160 on Quantitative is the 50th — a gap of 82 − 50 = 32 percentile points between two scores that look the same. That is why "what is a good GRE score" has no one-number answer, and why the GRE score percentile calculator asks which measure you mean before it tells you anything.

The rank is a definition, not a formula

No arithmetic converts a scaled score into a percentile. The rank is a straight table lookup:

percentile = ETS 2026–27 table[measure, scaled score] — the percentage of the reference group who scored lower than that score

The word lower is doing precise work. Some score-interpretation sources define percentiles as the share scoring "at or below" a value, which folds in everyone who tied with you; the ETS tables count only the people strictly beneath you. Take that Verbal 160: 82% of the reference group scored lower, so the remaining 100 − 82 = 18% scored 160 or above — and the rank alone cannot tell you how many of those 18 points are ties at exactly 160. On a scale with only 170 − 130 = 40 one-point steps — 41 possible scores — shared by a large testing population, ties at every score are guaranteed, which is why the definition behind a rank changes the number you should report.

Same score, different standing

Those three rows, all read straight from the ETS tables, carry the entire lesson. On Quantitative, the share of the group at 160 or above is 100 − 50 = 50%; on Verbal it is 100 − 82 = 18%. So roughly 50 ÷ 18 ≈ 2.8 times as many test takers reach 160-or-better on Quant as on Verbal — nearly three times the company at the same scaled number. A quantitative 160 sits at the group's midpoint; a verbal 160 sits above four-fifths of it. Each measure has its own distribution, and ETS publishes a separate percentile column for each one because the columns genuinely differ.

Analytical Writing compresses this even further. With scores running 0 to 6 in halves, there are only 6 ÷ 0.5 = 12 steps above zero — 13 possible scores in all — so a single half-point can move you past a substantial slice of the group. A 4.0 already places you above 63% of the reference group.

So what is a good GRE score?

The honest answer: a good score is a percentile that is competitive inside the applicant pool you are entering, on the measure that pool cares about. ETS itself cautions against overreliance on any single cut score and recommends that programs treat scores as one part of holistic review, alongside records, recommendations, experience and writing. There is no official threshold labeled "good."

The combined-score habit makes this harder to see, not easier. Two applicants can both report 160 + 160 = 320, yet the total hides which half did the work: that 320 is an 82nd-percentile Verbal performance stapled to a 50th-percentile Quantitative one. A program that weights quantitative preparation reads it very differently from one that weights verbal reasoning — which is exactly why score users are told to consider the measures separately. Before deciding whether to retest, translate each section score into its percentile (the lookup tool reads the current tables directly) and compare measure by measure against what your target department actually emphasizes.

Percentiles expire; scaled scores don't

A scaled score is equated, meaning a 160 earned on one test date represents the same performance level as a 160 from another. Percentile ranks have no such stability: they describe a particular reference group, and the group changes. The 2026–27 edition describes people who tested from July 1, 2022 through June 30, 2025; the next edition will describe a different population, and the same 160 can carry a different rank in it. Never weigh a percentile printed on an old score report against one looked up in a newer table — that compares two different crowds. And keep the edition attached to any rank you record, which is why the calculator prints its edition and reference period with every result.

The table's edges

• Blank cells stay blank. Where ETS reports no percentile information — the guide explains these are cells with no test takers above or below the specified scale-score range — the correct output is "Not reported," not zero and not an extrapolated guess.
• Half points only on Writing. A 4.5 exists; a 4.25 does not. Verbal and Quantitative accept whole points only.
• General Test only. Tables 1B and 1C say nothing about the Mathematics, Physics or Psychology Subject Tests, and they cannot convert a GRE score into a GMAT or LSAT equivalent.
• A percentile is not percent-correct. Verbal and Quantitative raw performance is equated onto the 130–170 scale before any percentile enters the picture; an 82nd-percentile score does not mean 82% of questions answered right.

The MCAT counts differently

If you are weighing graduate school against medical school, the MCAT's percentile convention is instructively different. The AAMC reports the percentage of examinees scoring at or below each score: ties count. A total score of 508 maps to 74 in the table effective May 1, 2026 through April 30, 2027, meaning 74% scored 508 or lower and 100 − 74 = 26% scored strictly above it. Compare the grammar: the GRE's 82 excludes your ties, the MCAT's 74 includes them. The two ranks are not the same kind of statement, even before you notice the populations differ.

The MCAT total spans 528 − 472 = 56 steps — 57 possible totals — while each of its four sections runs 118 to 132. The AAMC publishes its lowest totals, 472 through 474, as "less than 1%" rather than a decimal, and the MCAT score percentile calculator preserves that notation the same way the GRE tool preserves blank cells: the table's exact words, nothing invented. Section ranks live in their own columns too — a 125 on Critical Analysis and Reasoning Skills reads 62 — and neither adding nor averaging four section ranks reconstructs the total-score rank, which has its own column.

Reading your report, in order

When the report arrives: read the scaled scores first, because they are the durable part; look each one up in the current edition for its rank; note the edition; and only then start comparing against programs — measure by measure, never as a blended total. Both percentile tools live in the education section of Quanta, next to the GPA and grade calculators most applicants end up needing in the same season. One recording habit closes the loop: wherever a percentile lands in your notes, write three things beside it — the measure, the edition, and which definition of rank it uses, lower-than or at-or-below. A rank missing any of those three is a number without a sentence attached, and guessing the missing part is the one mistake a table lookup cannot protect you from. Our contact page exists for precisely that situation — a question about which convention a document is using costs a minute; a wrong guess can follow a score through an entire application cycle.

Sources

• ETS — GRE Guide to the Use of Scores 2026–27 (percentile tables)
• ETS — interpreting GRE scores
• AAMC — MCAT percentile ranks and their annual update
• AERA, APA & NCME — Standards for Educational and Psychological Testing