Early Decision Acceptance Rates at Top Colleges (CDS)
ED acceptance rates for 29 selective colleges from each school's own Common Data Set, with derived non-ED rates and what the gap does and doesn't mean.
Early Decision Acceptance Rates at Top Colleges, From Each School's Common Data Set
At most selective colleges that publish the numbers, the Early Decision admit rate is roughly double to triple the overall rate. That gap is real, but it does not show that applying ED would multiply your child's odds by the same amount.
The figures below come from each college's own Common Data Set (CDS). The CDS is a standard report that colleges fill in and post on their institutional-research sites. Section C21 asks how many ED applications a college received and how many it admitted. Section C1 gives the totals for all first-year applicants. We downloaded every file on September 24, 2026. Most are the 2025–26 CDS, which describes the class that entered in fall 2025.
Why Harvard, Yale, Princeton and Stanford Are Missing
Parents usually search for the Ivy League first, so start with what the CDS cannot tell you.
The CDS collects counts only for Early Decision. Section C22 (Early Action) asks whether a college offers early action, its dates, and whether the plan is restrictive. It asks for no application or admit counts. So no CDS-based early acceptance rate exists for colleges that run early action instead of ED:
- Harvard, Yale, Princeton, Stanford and Caltech each answer "No" to ED and "Yes" to a restrictive early action plan in their 2025–26 CDS.
- MIT offers non-restrictive early action and no ED.
- Duke and UChicago answer "Yes" to ED but leave the ED application and admit counts blank.
- Cornell, Tufts and several liberal arts colleges publish the counts, but label the line "Fall 2024" inside their 2025–26 file. We show their ED rate but do not combine it with fall 2025 totals.
When a website gives an "early acceptance rate" for Harvard or Yale, it comes from a press release, a campus newspaper, or an estimate. It does not come from the CDS. For how those plans work, see our guide to restrictive early action rules at Harvard, Yale, Princeton and Stanford.
The Table: ED Rate, Overall Rate, and the Rest of the Pool
How to read it. "ED" is admitted ÷ applied under C21. "All" is admitted ÷ applied under C1, and C1 includes ED applicants. "Non-ED (derived)" removes the ED applicants and admits from the C1 totals. Everything else stays in that group: regular decision, early action where offered, ED applicants deferred and admitted later, and waitlist admits. "ED share" is the share of all admitted students who were admitted through ED.
National universities
| College | CDS year | ED rate | All applicants | Non-ED (derived) | ED share of admits | Source |
|---|---|---|---|---|---|---|
| Columbia (College + Engineering) | 2025–26 | 819 / 5,868 = 14.0% | 2,946 / 59,611 = 4.9% | 4.0% | 28% | CDS |
| Penn | 2025–26 | 1,266 / 9,503 = 13.3% | 3,570 / 72,544 = 4.9% | 3.7% | 35% | CDS |
| Brown | 2025–26 | 907 / 5,062 = 17.9% | 2,710 / 42,774 = 6.3% | 4.8% | 33% | CDS |
| Dartmouth | 2025–26 | 768 / 3,552 = 21.6% | 1,699 / 28,230 = 6.0% | 3.8% | 45% | CDS |
| Vanderbilt | 2025–26 | 874 / 6,202 = 14.1% | 2,593 / 48,196 = 5.4% | 4.1% | 34% | CDS |
| Johns Hopkins | 2025–26 | 835 / 7,639 = 10.9% | 3,072 / 50,259 = 6.1% | 5.2% | 27% | CDS |
| Northwestern | 2025–26 | 1,212 / 5,939 = 20.4% | 3,933 / 52,676 = 7.5%¹ | 5.8% | 31% | CDS |
| Rice | 2025–26 | 647 / 5,748 = 11.3% | 2,948 / 36,791 = 8.0% | 7.4% | 22% | CDS |
| Cornell | 2025–26² | 1,889 / 10,057 = 18.8% | 6,077 / 72,523 = 8.4% | not derived | — | CDS |
| Emory | 2025–26 | 1,049 / 4,728 = 22.2% | 3,869 / 37,561 = 10.3% | 8.6% | 27% | CDS |
| Tufts | 2025–26² | 1,071 / 3,344 = 32.0% | 3,612 / 33,395 = 10.8% | not derived | — | CDS |
| Carnegie Mellon | 2025–26 | 553 / 2,680 = 20.6% | 3,859 / 34,867 = 11.1% | 10.3% | 14% | CDS |
| WashU | 2025–26 | 1,333 / 4,934 = 27.0% | 3,968 / 33,283 = 11.9% | 9.3% | 34% | CDS |
| Boston University | 2025–26 | 2,165 / 6,907 = 31.3% | 9,853 / 76,776 = 12.8% | 11.0% | 22% | CDS |
| Boston College | 2024–25 | 1,434 / 4,288 = 33.4% | 5,632 / 34,779 = 16.2% | 13.8% | 25% | CDS |
Liberal arts colleges
| College | CDS year | ED rate | All applicants | Non-ED (derived) | ED share of admits | Source |
|---|---|---|---|---|---|---|
| Pomona | 2023–24³ | 212 / 1,691 = 12.5% | 819 / 12,121 = 6.8% | 5.8% | 26% | CDS |
| Bowdoin | 2025–26² | 296 / 1,985 = 14.9% | 957 / 14,045 = 6.8% | not derived | — | CDS |
| Swarthmore | 2025–26 | 228 / 1,281 = 17.8% | 977 / 12,995 = 7.5% | 6.4% | 23% | CDS |
| Amherst | 2025–26² | 216 / 971 = 22.2% | 1,222 / 15,819 = 7.7% | not derived | — | CDS §C |
| Williams | 2025–26 | 256 / 964 = 26.6% | 1,336 / 15,520 = 8.6% | 7.4% | 19% | CDS |
| Claremont McKenna | 2025–26 | 216 / 978 = 22.1% | 655 / 6,939 = 9.4% | 7.4% | 33% | CDS |
| Barnard | 2025–26² | 476 / 1,437 = 33.1% | 1,048 / 10,426 = 10.1% | not derived | — | CDS |
| Harvey Mudd | 2025–26² | 109 / 594 = 18.4% | 643 / 5,217 = 12.3% | not derived | — | CDS |
| Davidson | 2025–26² | 358 / 1,232 = 29.1% | 1,127 / 8,932 = 12.6% | not derived | — | CDS |
| Hamilton | 2025–26 | 315 / 1,120 = 28.1% | 1,210 / 8,904 = 13.6% | 11.5% | 26% | CDS |
| Grinnell | 2024–25³ | 283 / 828 = 34.2% | 1,416 / 9,758 = 14.5% | 12.7% | 20% | CDS |
| Wellesley | 2025–26 | 301 / 1,023 = 29.4% | 1,258 / 8,506 = 14.8% | 12.8% | 24% | CDS |
| Vassar | 2025–26² | 402 / 1,046 = 38.4% | 2,327 / 11,118 = 20.9% | not derived | — | CDS |
| Carleton | 2025–26² | 263 / 760 = 34.6% | 1,480 / 6,796 = 21.8% | not derived | — | CDS |
¹ Northwestern leaves the printed C1 total blank. We added its printed rows by gender. ² The ED counts sit under a line reading "For the Fall 2024 entering class" inside the 2025–26 file. The ED rate is the college's own figure, but we do not mix it with fall 2025 totals. ³ The most recent CDS the college has posted as a downloadable file.
The arithmetic, once (Dartmouth, CDS 2025–26). All applicants: 1,699 ÷ 28,230 = 6.0%. ED: 768 ÷ 3,552 = 21.6%. Everyone else: (1,699 − 768) ÷ (28,230 − 3,552) = 931 ÷ 24,678 = 3.8%. So the ED round admitted at about 5.7 times the rate of the rest of the pool. ED admits were 45% of everyone Dartmouth admitted.
Left out on purpose. Duke and UChicago leave the counts blank. NYU's public file and UVA's dashboard gave us no ED counts we could verify. Middlebury reports only ED I admits ("excludes Feb admits"). Colby's site could not be reached. Wesleyan's newest public file is from 2022–23.
Why the ED Gap Overstates Your Child's Advantage
The numbers show that ED applicants were admitted at higher rates. They do not show that the ED choice caused the difference. Three points are worth keeping in mind before you treat the gap as a tactic.
1. The ED pool is not a random sample. Families who apply ED have to commit early, which usually means an early-ready application. Many are also comfortable binding themselves before comparing aid offers. For guidance on that trade-off, see our page on whether you can back out of Early Decision for financial aid.
2. Some ED admits may have a "hook" (hypothesis). A plausible explanation for part of the gap is that recruited athletes and, where legacy status is considered, alumni children apply disproportionately in early rounds. None of the CDS files above breaks ED admits down by athlete or legacy status, and we found no primary source that does for these colleges. Treat this as a plausible reason the gap looks larger for an unhooked applicant than it really is, not as a measured fact. The CDS does record whether a college considers alumni relations at all (item C7). We map that for more than 40 colleges in our legacy admissions status table.
3. Share of the class matters as much as rate. At Dartmouth and Penn, more than a third of admitted students came through ED. That leaves fewer seats for regular decision, which is part of why the derived non-ED rates above are lower than the headline rates.
The honest reading: ED is a real signal of commitment, and at some colleges it is rewarded. But an applicant who would be an average fit in regular decision is not transformed by applying ED. No table can tell you how a specific college will read one specific file.
A Decision Procedure for This Week
Run this before your teen signs any ED agreement.
- Confirm the plan type. Is it ED (binding), restrictive EA, or plain EA? Check the college's own page and its CDS C21/C22 answers. Our 2026–27 notification dates guide lists the timing.
- Run the net price calculator together. Do this before choosing ED, not after. ED commits your family before you can compare offers.
- Look up that college's row above. Compare the ED rate with the derived non-ED rate, not with the headline rate. Note the CDS year.
- Ask whether your teen is the typical ED applicant there. If the answer depends on an athletic recruitment or legacy status you do not have, discount the gap.
- Make sure the list works without the ED school. Our guide to how many colleges your child should apply to covers list shape.
- Plan for a deferral. Know what the college wants from deferred students. See what to do after an Early Decision deferral.
- Keep the essays your teen's own. The early round moves the essay deadline to early or mid-November at the colleges above (C21 closing dates). It does not change who writes.
The Part Still Under Your Teen's Control
By September, grades through junior year and most activities are set. The writing is not. Admissions offices vary in how much weight they give it. Our look at what 185 colleges said about essay importance shows the spread, and how holistic readers read a college essay explains what readers look for. If your teen is aiming at an early deadline, they can run their own draft through GradPilot's free college essay review against the Common App narrative craft rubric. Then read the feedback together. The review gives qualitative feedback on the draft; it does not predict admission, and your teen stays the author. The 2026–27 Common App prompts are a good place to start.
FAQ
Does applying Early Decision really increase my child's chances?
Every college in the tables above admitted ED applicants at a higher rate than its overall rate. That is an observed difference, not a promise. It is partly a difference in who applies ED. The CDS does not separate recruited athletes or legacies within ED, so the benefit to an unhooked applicant cannot be measured from these files.
Why can't I find Harvard's or Yale's early acceptance rate in the Common Data Set?
The CDS asks for application and admit counts only for Early Decision (C21). Early action (C22) asks for dates and whether the plan is restrictive, not counts. Harvard, Yale, Princeton and Stanford use restrictive early action, so their CDS contains no early admit rate.
Which CDS year should I trust?
Use the newest one the college has posted, and note which entering class it describes. The 2025–26 CDS covers students who entered in fall 2025. It says nothing directly about the fall 2027 class your teen may be applying for.
What does "non-ED (derived)" include?
It is everything in C1 minus the C21 ED figures. That includes regular decision, early action where offered, ED applicants deferred and later admitted, and waitlist admits. It is an approximation of "not admitted in the ED round," not a clean regular-decision rate.
Is Early Decision II counted in these numbers?
C21 asks for total early decision applications and admits. Where a college runs ED I and ED II, most report a combined figure. Middlebury is the exception we found: it reports ED I admits only, so we left it out.
Sources
All retrieved September 24, 2026.
- Columbia College and Columbia Engineering CDS 2025–26
- University of Pennsylvania CDS 2025–26
- Brown University Office of Institutional Research: Common Data Set
- Dartmouth CDS 2025–26
- Vanderbilt CDS 2025–26 (XLSX)
- Johns Hopkins CDS 2025–26
- Northwestern CDS 2025–26
- Rice CDS 2025–26
- Cornell CDS 2025–26
- Emory CDS 2025–26
- Tufts CDS 2025–26
- Carnegie Mellon CDS 2025–26
- WashU CDS 2025–26
- Boston University CDS 2025–26
- Boston College CDS 2024–25
- Pomona CDS 2023–24
- Bowdoin CDS 2025–26
- Swarthmore CDS 2025–26
- Amherst CDS 2025–26, Section C
- Williams CDS 2025–26
- Claremont McKenna CDS 2025–26
- Barnard CDS 2025–26
- Harvey Mudd CDS 2025–26
- Davidson CDS 2025–26
- Hamilton CDS 2025–26
- Grinnell CDS 2024–25
- Wellesley CDS 2025–26
- Vassar CDS 2025–26
- Carleton CDS 2025–26
- Yale CDS 2025–26 (C21/C22 answers and CDS definitions)
- Harvard CDS 2025–26
- Stanford Institutional Research: Common Data Set
- Princeton Institutional Research: Common Data Set
- MIT CDS 2025–26
- Caltech CDS 2025–26
- Duke CDS 2025–26
- UChicago CDS 2025–26
- Middlebury CDS 2025–26
Let Your Teen Check Their Own Draft
Your student runs their essay against the matching Common App rubric for free. Read the feedback together; the writing stays theirs.