What Is a Good SAT Maths Score?

The real Math distribution, not the round numbers

The SAT reports Math as a section score from 200 to 800. The College Board spells the section Math, American style, so that is how it appears on the report even though the subject is maths everywhere else on this site.

Here is what the class of 2025 actually did on it. The mean was 508 and the standard deviation 126, and the bands below account for the whole cohort of just over two million:

• 700-800 — 185,306 students, 9% of the cohort
• 600-690 — 317,263 students, 16%
• 500-590 — 484,027 students, 24%
• 400-490 — 643,389 students, 32%
• 300-390 — 329,868 students, 16%
• 200-290 — 45,112 students, about 2%

Before good, there is ready

There are two different questions hiding inside this one. Whether a Math score is competitive is a question about other applicants. Whether it is adequate is a question about the coursework it leads to, and the College Board answers that one directly with a benchmark of 530.

The benchmark represents a 75% likelihood of achieving a C or higher in a related first-semester credit-bearing college course. In the class of 2025, 41% of test takers met it — against 64% who met the Reading and Writing benchmark of 480. Math is where most of the shortfall sits, by a wide margin.

That asymmetry is the reason we treat the Math section as the first target in almost every SAT plan. It is where the largest group of students is furthest from where they need to be, and it is also the section that responds fastest to structured work.

Where the curve turns steep

Going from 480 to 560 and going from 700 to 780 are not the same task, and it is misleading to describe both as 80 points. In the middle of the range, points come from content — topics that were never secure become secure, and a whole class of questions starts going right.

Near the top, there is no content left to learn. A student at 700 is already answering most questions correctly, so further points come from eliminating error classes: the misread condition, the sign lost in transcription, the question abandoned too late. That is slower, less satisfying work, and it produces smaller gains for the same hours.

The evidence is consistent with this. In the largest causal study of SAT retaking, students in the lower half of the distribution gained roughly four times as much on a second sitting as those nearer the top, with the authors attributing part of the difference to ceiling effects. Anyone planning a push from an already strong Math score should budget accordingly.

Deciding what good means for your course

For a humanities or social science application, a Math score at or a little above the benchmark, paired with a strong Reading and Writing score, is a perfectly sound profile. Nobody is reading it in isolation.

For anything quantitative — engineering, computer science, economics, the physical sciences — the Math score is read on its own and the useful target is 700 or better. The 25% of test takers above 600 is a broad group; the 9% above 700 is not, and that is where the section stops being a liability.

The way to settle it for a particular student is the same as for the composite: find the section-level 25th and 75th percentile figures in section C9 of the Common Data Set for each university on the list, and aim above the 75th figure for the courses that will scrutinise it.