Every school's Common Data Set has a section built exactly for this -- Section C2 asks for the number of applicants offered a waitlist spot, the number who accepted it, and the number ultimately admitted from it. In theory, that means every Ivy League school's real waitlist odds should be a matter of public record. In practice, checking all 8 filings directly shows something different: only four of them disclose enough to actually compute a rate. Our full explainer on what "Ivy League" actually means covers other places these eight schools' disclosure habits diverge just as sharply.

The four that disclose enough to compute a real rate

Cornell reports the full breakdown for Fall 2024 (Class of 2028): 8,103 applicants offered a waitlist spot, 6,190 accepted it, and 388 were admitted -- a 6.3% admit rate among those who accepted, the highest of the four.

Yale reports the full breakdown for Fall 2025 (Class of 2029): 944 offered, 738 accepted, 32 admitted -- a 4.3% rate.

Princeton reports the full breakdown for Fall 2025 (Class of 2029): 1,370 offered, 1,086 accepted, 36 admitted -- a 3.3% rate. Princeton's own Common Data Set page blocks automated access, so this figure rests on two independent third-party compilations of that filing rather than a direct fetch -- both agree exactly, which is the same corroboration standard used elsewhere on this site when an official page can't be reached directly.

Penn reports accepted and admitted for Fall 2024 (Class of 2028) -- 2,288 accepted a spot, 66 admitted, a 2.9% rate -- but its offered figure for that specific cycle isn't confirmed here, so only the two-figure ratio is used.

The three that disclose only part of it

Harvard's Fall 2025 (Class of 2029) filing reports only the number of students admitted from its waitlist -- 75 -- and leaves both the offered and accepted fields blank. There's no way to turn that into a rate; 75 admits could be a strong showing off a small waitlist or a thin one off a huge pool, and the filing doesn't say which.

Dartmouth does the opposite: its Fall 2025 (Class of 2029) filing reports 2,497 applicants offered a waitlist spot, then leaves both the accepted and admitted fields blank. Dartmouth's prior cycle (Class of 2028, Fall 2024) did report all three -- 2,589 offered, 2,189 accepted, 29 admitted -- so this isn't a standing policy of withholding the data, just a gap in this specific year's filing.

Brown doesn't publish the offered/accepted breakdown in a form confirmed here, but a reported 118 students were admitted off its waitlist for Fall 2024 (Class of 2028) -- up from 73 the year before, out of an entering class of roughly 1,700.

The one that discloses nothing

Columbia's own Fall 2024 (Class of 2028) Common Data Set confirms, with a checked box, that it does place students on a waitlist -- then reports none of the three numbers at all. The form simply moves on to the next question. Of the 8 Ivies, Columbia is the only one where the underlying policy is confirmed but every figure describing it is withheld.

Why this is worth checking yourself, not trusting a cached number

Waitlist admit rates swing hard year to year at the same school, driven by how that year's yield came in -- our general guide to waitlist strategy covers why in more depth, using Michigan's own admit rate nearly doubling between two consecutive fall cycles as the clearest example. Ivy League waitlists are no exception: Dartmouth's own two most recent filings show admitted figures of 29 and then blank, not a stable number either. Whatever rate you find here or anywhere else describes a specific school in a specific year -- not a number worth carrying forward to your own cycle.

What this means for you

  • Only 4 of the 8 Ivies -- Cornell, Yale, Princeton, and Penn -- currently disclose enough in their own Common Data Set to compute a real waitlist admit rate.
  • Harvard, Dartmouth, and Brown each disclose a piece of the picture (an admitted count, an offered count, or a reported admitted figure from secondary reporting) but not the full funnel.
  • Columbia confirms it uses a waitlist but discloses none of the three underlying numbers.
  • None of these numbers predict your own odds -- check the specific school's current Common Data Set Section C2 when you're actually on a waitlist, and see what to actually do once you're there.
  • This is also the same document behind what each Ivy actually weighs in admissions and what's disclosed about legacy admissions -- one filing, asked every year, and each school chooses what to fill in.
  • The same pattern shows up outside the Common Data Set, too -- checking each Ivy's own newspaper coverage of its early vs. regular decision round found only 2 of 8 currently disclose a real split there either.

Sources