How the GRE Is Scored

The three GRE scores and what each scale means

The GRE produces three independent scores, not one number. Verbal Reasoning is reported on a 130-170 scale in 1-point increments. Quantitative Reasoning uses the same 130-170 scale. Analytical Writing is reported separately on a 0-6 scale in half-point increments. There is no official combined total; when people quote a single GRE score around 320, they are informally adding Verbal and Quant, which together span 260 to 340.

Because the scores are separate, programs weigh them differently. A literature, law-adjacent, or social-science department may care most about your Verbal number, while engineering and economics programs lean on Quant. This matters for how you study: if your target programs emphasize Verbal, the highest-leverage work is closing the gap on the question types and vocabulary you keep missing, not spreading effort evenly.

Alongside each scaled score, ETS reports a percentile rank that tells schools what share of recent test-takers you outscored. Percentiles are recalculated periodically against a rolling pool of examinees, so the same scaled score can map to slightly different percentiles over time. The scaled score is what you control; the percentile is context.

Section-adaptive scoring: how the test adjusts to you

The GRE is section-adaptive, which is different from question-adaptive tests like the GMAT. Verbal Reasoning consists of two sections totaling 27 questions in about 41 minutes. The first section is always medium difficulty and is the same starting point for everyone. Your performance on that first section determines whether your second section is delivered at an easier, medium, or harder difficulty tier.

This design has a direct strategic consequence: the first section matters more than its raw question count suggests, because it sets your ceiling. Reaching the hard tier in the second section is what makes the top of the 130-170 range reachable, while landing in an easier tier caps how high your scaled score can go. There is no penalty structure that punishes you for the adaptive routing itself; you are simply being measured against a harder or easier set of questions in section two.

Quantitative Reasoning works the same way: two sections, 27 questions, about 47 minutes, with section two's difficulty set by section one. Within any single section you can move freely, skip, flag, and revisit questions before time runs out, so the adaptation happens between sections, never mid-section. That means pacing and not stalling on the first section is as important as accuracy.

From raw score to scaled score: equating

Your raw score is simply the number of questions you answer correctly in each measure; there is no penalty for wrong answers, so you should answer every question even if you have to guess. ETS converts that raw count into the 130-170 scaled score through a statistical process called equating.

Equating does two jobs at once. First, it accounts for the difficulty tier you were routed into, so a correct answer on a hard second section is worth more toward your scaled score than a correct answer on an easier one. Second, it adjusts for small difficulty differences between the many editions of the test, so that a 162 means the same level of ability regardless of which version you happened to sit. This is why there is no fixed, public raw-to-scaled conversion table; ETS does not publish the exact equating formula, and it varies by edition.

A practical takeaway: do not try to reverse-engineer your score from a question count during the test. The honest preparation move is to raise your true ability on the underlying skills, because equating rewards accuracy on harder material. For Verbal, that means mastering Text Completion and Sentence Equivalence answer mechanics, building a strong high-frequency vocabulary base, and reading carefully for Reading Comprehension.

How multi-blank and Sentence Equivalence questions are scored

Verbal scoring has two question types where partial credit does not exist, and understanding this changes how you allocate effort. Text Completion questions can have one, two, or three blanks. For two- and three-blank questions, you must select the correct word for every blank to earn the point; getting two of three blanks right scores the same as getting zero. There is no partial credit.

Sentence Equivalence asks you to pick two of six answer choices that both fit the sentence and produce sentences alike in meaning. You only earn the point if both of your selections are correct. Choosing one right and one wrong scores zero, just like a two-blank Text Completion with one error.

This all-or-nothing structure is why method matters more than guessing. The reliable approach is to predict the meaning the blank needs before reading the choices, then use signal words and connotation to eliminate, and for Sentence Equivalence to actively look for the matched pair rather than two words you simply like. Grezi's 19 interactive strategy lessons drill exactly these mechanics, and its adaptive practice resurfaces the blank patterns and confusable words you keep missing.

How Analytical Writing is scored, and how to read the rubric

Since the 2023 redesign, Analytical Writing is a single 30-minute task: Analyze an Issue. You read a brief claim and write an essay taking and defending a position. It is scored holistically from 0 to 6 in half-point steps. Scoring uses two graders in parallel: a trained human reader and an automated engine called the e-rater.

If the human score and the e-rater score agree closely, their average becomes your reported score. If they diverge by more than a small tolerance, a second trained human reads the essay, and the final score is the average of the two human scores. The e-rater is a check on consistency, not the sole judge, so essays are not graded purely by machine.

The rubric rewards a clearly stated position, well-developed and relevant reasons and examples, logical organization, and controlled, varied language; surface grammar matters less than the strength of the argument. Practically, that means plan a thesis before you write, support it with two or three concrete reasons, address a counterpoint, and leave time to proofread. AWA is reported separately and does not affect your Verbal or Quant scaled scores.

Score reporting, ScoreSelect, and validity

GRE scores are valid for five years from your test date, and they are typically available in your ETS account about 8 to 10 days after a computer-based test. Your official report shows your Verbal, Quant, and Analytical Writing scores along with their percentile ranks.

ScoreSelect is the feature that gives you control over what schools see. On test day you can send scores to recipients for free using either your most recent test date or all dates from the past five years. After test day, for a fee, you can send scores from any single test date, your most recent, or all dates within the five-year window. Schools receive only the dates you choose, and on any date you send, they see all three section scores from that sitting; you cannot mix and match individual sections across different dates.

Because ScoreSelect lets you suppress a weaker attempt, a disappointing first sitting does not have to follow you. That lowers the cost of taking the test before you are perfectly ready, but it also means your preparation should target a confident single-day performance. Knowing the score ranges your target programs expect, then working backward to the Verbal and Quant levels you need, turns the abstract 130-170 scale into a concrete study target.

Frequently asked questions