Ivy League SAT scores: published ranges and setting a target

01Read the cohort before comparing the numbers

The table uses primary university sources checked for this update. The years and populations differ. An admitted-student range includes people who may enroll elsewhere; an enrolled-student range describes the students who actually joined the class. Both cover score submitters rather than everyone who applied.

Harvard's published 2024–25 Common Data Set supplies section percentiles but leaves the SAT total row blank. Its section ranges are shown directly below. Adding section percentiles would not produce a valid total-score percentile, since the students at each section percentile need not be the same students.

02The 25th percentile is not a minimum

For a reported range of 1490–1560, the low endpoint is the 25th percentile and the high endpoint is the 75th percentile in the described group. Roughly a quarter have scores at or below the low endpoint and roughly a quarter at or above the high one; ties can make the exact proportions less tidy.

A 1490 therefore isn't a rejection threshold. A 1560 isn't an acceptance threshold. The table doesn't include the scores and full academic records of rejected applicants, so it cannot tell you your admission probability.

Be especially careful with data from test-optional cohorts. Students who chose to submit scores can differ from students who didn't submit them. That selection affects what the published range describes.

03Set a useful score goal for your own list

Use the 75th percentile as a possible preparation target if it is realistic for your timeline. It's a way to choose a goal, not a promise that crossing it changes your admission outcome by a known amount.

Suppose your recent timed results are 1460–1500 and your selected schools publish upper endpoints around 1550–1560. A 1550 target gives you a concrete gap to investigate. Break it into sections: a 750 R&W and 710 Math result calls for different practice from a 690 R&W and 770 Math result, even though both total 1460.

Compare the published section ranges when available, especially if the program you want depends heavily on one subject. Keep grades, course choices, and application deadlines in the same planning conversation. I wouldn't turn every spare week into another test attempt solely because one school reports a slightly higher percentile.

04Match the testing policy to your application cycle

For the 2026–27 application cycle, Princeton remains test optional. It says it will require testing beginning with the 2027–28 cycle. Columbia College and Columbia Engineering announced the same cycle change: optional in 2026–27, required beginning in August 2027 for the 2027–28 cycle. See Princeton's policy and Columbia's announcement.

For every school, record the testing rule for the year you will apply, accepted alternatives or exceptions, score-reporting method, and latest accepted test. A class profile from a former optional year does not establish today's requirement. School-specific instructions also decide whether you may self-report scores and how superscoring works.

05Decide whether a retake has a specific job

A retake makes more sense when your practice review identifies work you can complete before the next accepted date. Examples include punctuation questions you can now explain, algebra setups you keep misreading, or a pacing issue that leaves several questions unanswered.

Take a fresh practice test, review every missed and guessed question, then choose focused work in the bank. Use several timed results rather than one favorable estimate to judge readiness. Don't use an estimated score calculator to predict the exact official result from a fixed number of misses.

If you already meet your own score goal and haven't found a clear reason to retake, put the time into the rest of your application. Keep a balanced college list; a high SAT score alone doesn't turn an Ivy into a safety.