SAT Score for Ivy League: Real Targets for Fall 2027

17 min read

Six of the eight Ivy League schools require an SAT or ACT score from students applying this fall for fall 2027 entry. For those six, the working target is a composite at or above your hardest target's published 25th percentile, which sits between roughly 1470 and 1510 across the league. A 1550 closes the question. Everything below is the band table, the current policies, and a framework for deciding whether to test again.

The short answer: the SAT score that makes you competitive at each Ivy

The honest version of the answer has three tiers, and almost every student applying to the eight Ivy League schools falls into one of them.

1550 and above: your score is a settled asset. It sits at or near the 75th percentile at every Ivy. No additional point you earn changes an admissions decision.

1500 to 1549: you are inside the published band at most of the league. The score keeps you in the pool. It does not distinguish you, because thousands of other applicants are standing in the same place.

1470 to 1499: you are around the 25th percentile at several Ivies. The score is usable, especially in context (more on that below), but something else in the file has to do real work.

Below about 1470: at the six Ivies that now require testing, the score is a drag you need to offset with transcript rigor, an unusual activity record, or school context that explains it.

All eight Ivies: middle 50% SAT bands

These are composite SAT ranges for enrolled first-year students as reported in each school's 2024-25 Common Data Set, the class that entered in fall 2024. The midpoint column is simple arithmetic on the published boundaries, not a reported median, and it is there so you can see the practical center of each band.

Sources: institutional Common Data Set filings, including Harvard's Office of Institutional Research. Bands shift 10 to 20 points between filing years, so treat the linked filing as the authority. The decisions in this guide do not change if your target's 25th percentile moves by a handful of points. If you have never read one of these filings before, our walkthrough of how to decode a Common Data Set shows you which sections actually matter.

The thing almost nobody tells you about these numbers

A published band is a submitter band, not an applicant band, and not even an admit band. It describes enrolled students who chose to send a score. During the test-optional years that produced the 2024-25 filings, students with lower scores simply withheld them, which removed the bottom of the distribution from the published range.

So when you compare your 1480 against a band that starts at 1500, you are comparing yourself to a filtered group. The real admitted pool contained scores below that floor. You just cannot see them in the data.

Two practical consequences. First, stop treating the 25th percentile as a cutoff; it is a visibility threshold. Second, expect these bands to drift down over the next two filing cycles as required-testing classes replace test-optional ones.

Do you even have to send a score? Policies for fall 2027 vs fall 2028 entry

For students applying during the 2026-27 cycle (fall 2027 entry), six Ivies require an SAT or ACT: Brown, Cornell, Dartmouth, Harvard, Penn and Yale. Columbia and Princeton remain test-optional for this cycle, and both have announced score requirements beginning with the 2027-28 cycle, meaning fall 2028 entry.

Required now (fall 2027 entry)

Brown, Dartmouth and Harvard reinstated first, followed by Cornell and Penn. Yale's standardized testing policy is the one families most often get wrong: Yale spent two cycles under a "test-flexible" rule that let AP or IB exam scores stand in for the SAT or ACT. That substitution ended. For fall 2027 entry, Yale expects an SAT or ACT score. Several score-comparison pages still print "test-flexible" next to Yale. They are out of date.

Test-optional for one more cycle

Columbia's testing policy and Princeton's standardized testing page both remain optional for applicants this fall, with the requirement arriving for the following cycle. Columbia's announced requirement includes a hardship carve-out for applicants who cannot reasonably access a testing center, which is a narrow accommodation, not a general opt-out. If you intend to use it, document the barrier in your application rather than assuming it applies.

If you are a sophomore or junior right now

Plan as if all eight require a score. A student entering college in fall 2028 or later has no Ivy left that is reliably test-optional, and building a four-school list around a shrinking optional policy is a bad bet. Register for a spring junior-year sitting and keep the summer free for one retake.

How admissions offices actually read your score

Your score is not measured against a national bar. It is measured against your high school.

When Dartmouth reinstated testing, it published the faculty analysis behind the decision. The economists who ran it found that scores helped the office identify high-achieving applicants from less-advantaged backgrounds whose transcripts were difficult to interpret on their own. Dartmouth's announcement of the research reported that among less-advantaged applicants scoring 1420 or above, submitting the score raised the probability of admission by about 4.3 percentage points, roughly two and a half times the baseline rate for that group.

Read that number carefully. A 1420 is below every Ivy's published 25th percentile. It was still a net positive for those students, because the office read it against what their school typically produced.

The mirror image

If you attend a high-scoring suburban public, a US boarding school or a competitive international prep school, the same logic runs the other way. A 1480 from a school where the top decile clears 1540 reads as mid-pack. The score is not bad. It is simply not information the reader did not already have.

This is why "is a 1450 good?" has no universal answer, and why chance-me threads are useless. The useful question is: is my score above or below what a reader expects from someone at my school with my transcript? Your counselor's school profile, which goes out with every transcript, contains the answer. Ask to see it.

When a below-band score is still an asset

• Your school sends few students to selective colleges and your score is far above its norm.
• You took the test cold, with no paid prep, and your transcript shows the same upward trajectory.
• Your section scores align with your intended field (a 790 Math from a school with no calculus offering is a signal).

The submit / retake / reallocate framework

Most students asking "should I retake the SAT for Ivy League admission" are really asking "where does my next 30 hours do the most good?" Four questions answer it.

Question 1. Is your composite above the 25th percentile of your hardest target? Use the table above. Yes or no.

Question 2. Is your composite above the top-decile score at your own high school? Ask your counselor for the school profile. Yes or no.

Question 3. Do you have a real path to plus 40 or more points? Not a wish, evidence. Take one full timed practice test under real conditions. College Board's own retesting research finds that most juniors who retest as seniors improve, with typical gains around 40 points. A 40-point gain moves you one band position at most. It is worth having only if you are sitting on a boundary.

Question 4. What does the next 30 hours buy? Thirty hours of prep at this level is worth perhaps 20 to 40 points. Thirty hours on a Common App essay is worth the difference between a readable file and a memorable one. The exchange rate is not close once you are in band.

The decision table

The reallocation rule

Score chasing becomes the most expensive mistake in an application the moment you are inside the band. A fourth or fifth sitting costs a registration fee, two weekends of prep and, more importantly, the October and November hours that your supplemental essays need. Students who run out of time in late October almost never ran out of points. They ran out of writing days.

Score by score: is 1400, 1450, 1500, 1550 or 1600 enough?

For national context: College Board's program results put the recent graduating-class SAT mean near 1030 across more than two million test takers, so every number in this section is already in the top few percent of American test takers. The SAT Suite program results publish the full distribution. Note that the percentiles printed on your score report are user percentiles, calculated against actual test takers, not against all US students. The nationally representative percentile is always higher and is not what colleges see.

1400 to 1449

Roughly the 93rd user percentile, and below every Ivy's published 25th percentile. It works when your school profile shows you well above local norms, when your transcript is the strongest in your graduating class, or when an activity record carries genuine distinction. At Columbia and Princeton this cycle, you can simply withhold it.

1450 to 1499

The ambiguity zone, and the one where school context decides everything. At a school where the top decile sits at 1420, this is a plus. At a feeder prep school, it is noise. One retake is defensible here if a practice test backs it.

1500 to 1549

In band at essentially every Ivy, around the 97th user percentile. This is the point where more prep stops paying. The score has done its job: it has confirmed that you can handle the academic work. Nothing about a 1530 versus a 1510 moves a committee. Close the prep book and go write.

1550 to 1600

The score question is closed. What decides the outcome now is the rest of the file: curriculum rigor, whether your extracurriculars show real depth, recommendation quality and essays. A 1600 does not guarantee anything, and tens of thousands of applicants with scores in this range are declined every year by schools with single-digit acceptance rates.

Section splits matter more than people think

A 1500 made of 760 Math and 740 EBRW reads differently from 740 Math and 760 EBRW if you are applying to Penn Engineering, Cornell Engineering or a quantitative program. Readers in those divisions look at the math section specifically. A humanities or social science applicant with a slightly lower math score and an outstanding EBRW is making a consistent case. Fix the section that contradicts your intended major, not the composite.

Why the Ivy bands look higher than they used to

Three forces pushed the published numbers up, and only one of them has anything to do with students getting smarter.

Self-selection during test-optional years. When submitting is a choice, only strong scorers submit. The published middle 50% of a test-optional class is mechanically higher than the same class would have reported under a mandate. As the required-testing classes filter into the Common Data Set over the next two filings, expect the 25th percentiles to settle lower.

Smaller, more self-selected applicant pools after reinstatement. Requiring a test discourages some applicants from applying at all. Reporting from The Harvard Crimson tracked a meaningful contraction in Harvard's applicant pool in its first required cycle. Fewer applicants with scores, higher reported bands.

Cornell's structure, which explains the league's lowest floor. Cornell admits to eight separate undergraduate colleges with genuinely different academic profiles, from Engineering to Agriculture and Life Sciences to Industrial and Labor Relations. The university-wide band is an average of eight different admissions conversations, which is why its 25th percentile sits below the rest of the league. If you are applying to Cornell, the college-specific requirements and essay structure matter far more than the university-wide number. Penn behaves the same way internally: Wharton and Nursing do not read a math score identically.

Which is why "the average SAT for the Ivy League" is a statistic with no decision value. It is eight distributions, several of them internally split, flattened into one number that describes no actual applicant.

Where Unive fits once the score question is settled

If your score is in band, the bottleneck in your application is no longer points. It is time, structure and feedback on the parts of the file that cannot be scored by a machine.

Unive is an AI admissions workspace built by Yale graduates that covers the parts of the application a test score cannot touch: a Profile Evaluator that reads your activities list the way a committee does, an Essay Grader with line-level feedback, an Application Roadmap that holds your deadlines in one place, scholarship matching, and interview prep with a voice agent. Students working inside it see a 3.48x higher acceptance rate. It is a paid subscription with a 7-day money-back guarantee, which is a different category of commitment from the $20,000 private counselor many families are comparing it against. If you are weighing that choice, we broke down what college consulting actually costs in detail.

What a strong score cannot do

Open section C7 of any Ivy's Common Data Set. That is the table where the school rates each admission factor as very important, important, considered or not considered. At most of the league, rigor of secondary school record and academic GPA sit in the top tier. Standardized test scores frequently sit a notch below, alongside essays, recommendations and character.

That ranking is not marketing language. It is a disclosure the school files annually. It tells you that a 1570 paired with a schedule that avoided the hardest courses available reads worse than a 1510 paired with the most demanding transcript your school permits. Our breakdown of what GPA Harvard admits actually have shows how tightly that top tier clusters.

What separates thousands of identical 1550s:

• Essays. The one place where a reader hears you rather than measures you. Start with the prompts and worked examples in our guide to the best Common App essays, and know the essay mistakes that quietly damage an application.
• Recommendations. Teachers who can describe how you think, not that you were pleasant.
• Activities with depth. Use a real standard rather than a count; our six-signal scoring rubric for activities gives you one.
• Fit with the specific school, which is what the supplements test.

One side benefit worth knowing: outside the Ivies, which meet need without merit awards, a high score is often directly convertible to money. Plenty of excellent universities attach automatic merit scholarships to score thresholds, which is a strong argument for a retake that has nothing to do with the Ivy League. Several schools on our list of hidden gem colleges fall into exactly that category.

Your testing calendar, counted backward from the fall 2027 deadlines

If you are a senior right now, your remaining calendar is short and the arithmetic is simple.

November 1, 2026 is the Early Decision and Early Action deadline across most of the league, including the restrictive early action programs at Harvard, Princeton and Yale. The November SAT sitting is generally the last one that lands in time for an early round, and most offices say so on their testing pages. Confirm the exact date on College Board's test dates and deadlines calendar before you count on it. If you are still deciding between rounds, our comparison of priority deadlines versus regular deadlines explains what each one actually commits you to.

Early January 2027 is when Regular Decision closes across the Ivies, clustering between January 1 and January 5. The December sitting is your last usable one, and it is genuinely last: digital SAT scores take roughly two weeks to release.

Superscoring and score choice. The Ivies generally consider your highest section scores across sittings, which is why a targeted retake aimed at one section can be worth more than a general one. Score-reporting rules differ, though: some schools ask you to report every sitting rather than pick. Check the testing page of each school on your list before you decide what to send.

Rising juniors aiming at fall 2028 entry: test in spring of junior year, retake once in late summer or fall of senior year, and stop. That plan keeps Columbia and Princeton open, since both require scores from your cycle forward.

Build the rest of the application in one place

A score is one line in a file that runs to dozens of fields, seven essay prompts and a dozen deadlines. Unive's Application Roadmap turns that into a sequenced plan with dates attached, the College Matching tool builds a reach, target and safety list grounded in real admissions data rather than reputation, and students who want a human in the loop can book 1-on-1 mentoring with graduates of top schools. Start with the AI admissions platform at unive.ai and let the test score be the one thing you have already finished.

SAT score for Ivy League FAQ

Which Ivy has the lowest SAT requirement?

None of the eight publishes a minimum score. Cornell's published 25th percentile is typically the lowest in the league, around 1470 in its 2024-25 Common Data Set, largely because Cornell admits to eight undergraduate colleges with different academic profiles. A low university-wide floor does not mean a low bar inside Engineering or Computer Science.

Can I get into an Ivy with a 1400?

Yes, and students do every year, but the score has to be read in context rather than against the published band. Dartmouth's own research found that among less-advantaged applicants scoring 1420 or above, submitting raised admit probability by about 4.3 percentage points. A 1400 from a high-scoring prep school carries far less weight than the same score from a school that rarely produces one.

Do the Ivies superscore the SAT?

Generally yes: Ivy admissions offices consider your highest section scores across multiple sittings when calculating your best composite. Score-reporting rules differ, though. Some schools expect you to report every sitting rather than choose. Check each school's standardized testing page before deciding what to send, because the policy, not the practice, governs what you must disclose.

Is the ACT easier for Ivy admission than the SAT?

No. Every Ivy accepts the SAT and ACT equally and converts between them using concordance tables. Switching is worth considering only if the test formats suit you differently: the ACT runs faster with a science reasoning section, the digital SAT is adaptive and shorter. Take one timed practice of each and switch only if the gap is clearly meaningful.

How many times should I take the SAT?

Two sittings is the standard plan, three is the practical ceiling. College Board's retesting research shows most juniors who retest as seniors improve, by roughly 40 points on average. Once you are inside your target school's published band, a fourth sitting costs you October and November writing time and buys almost nothing a committee will notice.

Does a perfect 1600 guarantee anything?

No. A 1600 removes the score as a question and nothing more. Ivy acceptance rates sit in the low single digits, and every school declines applicants with perfect scores each year. Curriculum rigor, GPA, essays, recommendations and the depth of your activities decide outcomes once your score is in band.

Are Columbia and Princeton still test-optional for me?

For fall 2027 entry (applying during the 2026-27 cycle), Columbia and Princeton remain test-optional, and both have announced requirements beginning with the following cycle, fall 2028 entry. Columbia's policy includes a narrow hardship waiver. If you are currently a junior or younger, plan on all eight Ivies requiring an SAT or ACT score.