What SAT Score Do You Need for the Ivy League?
The middle-50% SAT score ranges for every Ivy League school, plus how to think about the bar if you're aiming for admission.
The Middle 50% at Each Ivy
Middle-50% SAT ranges across the Ivies cluster between 1470 and 1570. That means 25% of admitted students score below the low number and 25% score above the high one.
- Harvard: ~1500–1580
- Princeton: ~1500–1570
- Yale: ~1500–1580
- Columbia: ~1490–1570
- UPenn: ~1490–1560
- Brown: ~1490–1560
- Dartmouth: ~1470–1560
- Cornell: ~1460–1550
What the Numbers Actually Mean
Hitting a school's 75th percentile SAT score doesn't guarantee admission — holistic review matters more than any single number. But scoring above that line takes standardized testing off the board as a weakness.
What to Aim For
If an Ivy is on your list, aim for 1500+. If you're applying to more than one, aim for the 75th percentile of the highest-range school you're serious about.
What to do after reading this
- Record your latest Reading and Writing score, Math score, and total score.
- Identify the section that is furthest from your target.
- Run one targeted drill in that section before taking another full module.
Related SAT guides
- What Is a Good SAT Score in 2026?
A breakdown of what counts as a good Digital SAT score, with percentiles, target colleges, and how the bar shifts depending on your goals.
- Average SAT Scores by College: How to Read the Ranges
How to use middle-50% SAT ranges from the Common Data Set to set realistic targets for any college on your list.
Mistakes to avoid
- Chasing a total-score goal without knowing which section is holding it down.
- Comparing scores without checking whether they are from timed, realistic practice.
- Taking another practice test before reviewing every missed question.
Use this in practice
Practice on 1600.now
Apply what you just read. Run the numbers in the SAT score calculator, take a full Digital SAT module, or drill targeted skills in the question bank.
Post 48 of 52 in the 1600.now Digital SAT blog.