At a glance
MIT admitted 4.58% of applicants to the Class of 2030. A close read of the round-by-round data, the test scores the Institute requires, the aid policy that changed who applies, and why legacy, donor ties and demonstrated interest carry no weight.
Every December, the same scene repeats in school counselling offices. A student sits with a laptop open at 6:28 on a Tuesday evening, refreshing a portal that has just gone live, while a parent hovers behind the chair. This year the refresh returned 655 offers of admission from 11,883 Early Action applications to the Massachusetts Institute of Technology, a 5.5% admit rate. Another 7,738 students were deferred and 2,703 were turned down outright.
By the time the regular round closed, 28,349 students had applied for the Class of 2030 and 1,299 had been admitted. That's 4.58%, the third cycle in a row inside a band between 4.5% and 4.6%. The headline figure gets quoted everywhere, and it is accurate. It also hides nearly everything a family needs in order to decide what to do next.
The 4.6% is an average of two different rounds
Start with the arithmetic for the Class of 2029, the most recent cycle MIT has published in complete detail. These are the Institute's own figures rather than a consultant's reconstruction.
| Applicant group | Applied | Admitted | Rate |
|---|---|---|---|
| All first-year applicants | 29,281 | 1,334 | 4.6% |
| Early Action | 12,052 | 721 | 6.0% |
| Regular Action, new applicants | 17,229 | 603 | 3.5% |
| U.S. citizens and permanent residents | 22,355 | 1,198 | 5.4% |
| International applicants | 6,926 | 136 | 2.0% |
| Wait list | 561 offered a place | 10 | 1.8% |
Three numbers matter more than the rest. The gap between rounds is real: 6.0% early, 3.5% regular. The international line is where the door narrows hardest, with 6,926 applicants competing for 136 places, against 5.4% for domestic applicants. And the published total only reconciles when you add the wait list: 721 early admits plus 603 regular admits makes 1,324, and the last ten came from a pool of 561 wait-listed students.
Then there is the deferral pool. MIT pushed 7,486 early applicants into the regular round instead of rejecting them, and 175 of those were later admitted. That works out at 2.3%, which means a deferral is a live path rather than a courtesy.
MIT is equally direct that applying early earns nothing by itself. In a 2022 post explaining its testing policy, Dean of Admissions Stu Schmill wrote that the Institute gives no “advantage to those who apply through our (non-binding and non-restrictive) Early Action process”. The higher early rate, on that account, reflects a pool that already had its testing, essays and college list finished by November.
What MIT actually asks for, in print
Testing comes first, because MIT is the outlier among its peers. The Institute dropped its SAT or ACT requirement for two years during the pandemic, then became the first prominent university to reinstate it, in March 2022, after its admissions research team went back through roughly fifteen years of records. The finding, as Schmill described it, was that scores sharpened predictions about who would finish an MIT degree, and that the pattern held across socioeconomic groups.
His explanation for why the maths score in particular carries weight is unusually concrete: every MIT undergraduate, whatever the eventual major, takes two semesters of calculus and two of calculus-based physics. “Our research can't explain why these tests are so predictive of academic preparedness for MIT, but we believe it is likely related to the centrality of mathematics—and mathematics examinations—in our education,” he wrote.
The published ranges for the Class of 2029 sit where you would expect them to. The middle 50% of admitted students scored 780 to 800 on SAT Math, 740 to 780 on SAT Evidence-Based Reading and Writing, and 34 to 36 on the ACT composite. What MIT does not publish is a cutoff. There is no stated minimum GPA and no score below which a file is thrown out, which is why those ranges describe an admitted class rather than a gate. “We do not prefer people with perfect scores,” Schmill said when the requirement came back.
The rest of the file is what you would expect: a school report, transcripts, teacher recommendations and essays, plus the February Updates and Notes Form that MIT sends to applicants it has deferred. Nothing on that list is secret. MIT's admissions office says plainly that it avoids “unnecessary complexities or hidden preferences”, and the published record backs the claim up.
The backdoors that don't exist
Ask about legacy and the answer runs to one sentence. MIT's own FAQ states that the Institute does not consider legacy or alumni relations, and its 2025–26 Common Data Set marks both alumni relation and level of applicant's interest as Not Considered. MIT's news office has been just as blunt: unlike many other institutions, it does not give an admissions advantage to the children of alumni or donors.
Demonstrated interest is the other half of that. Campus visits, information sessions, emails to regional admissions staff: none of it is tracked, so none of it moves a decision. Neither does timing, according to Schmill's 2022 post, which says in the same paragraph that MIT gives no preference to legacies and no weight to demonstrated interest. Families spend real money on all three anyway.
The strongest defence against a manufactured application is the context test. “We don't care as much about what a student has done as what they have done relative to what might have been expected, given their resources,” Schmill wrote. A summer programme bought at full price and a year of supervised lab work at the nearest public university are not the same claim on an application, and the readers are instructed to weigh each against the applicant's circumstances.
So what do we tell the family that arrives with a consultant's invoice and a list of alumni to call? The same thing the statistics page says. There is no lever, and the money is better spent on the parts of the process that are genuinely scarce: strong maths teaching, a test sitting that isn't rushed, and teachers who know the student well enough to write something true.
The narrow doors that are real
Transfer is the route families ask about least and should understand best. MIT's Common Data Set for 2024–25 records 1,346 transfer applicants and 32 admits, a rate of 2.38%, and recent years have ranged from roughly 1.3% to 3.1%. The mechanics are different from the first-year round: MIT runs its own transfer application rather than the Common App, the fall deadline lands in mid-March, applicants need at least a year of college behind them, and the two required recommendations should include a maths or science instructor. Test scores are required here too. The Institute's published statistics make the freshman numbers look generous by comparison.
Those odds are narrow because MIT graduates nearly everyone it admits, so few seats ever open. The other documented routes are narrower still.
- QuestBridge, the nonprofit that matches low-income students with selective colleges, sent 100 students into the Class of 2029, up from 56 the year before, a figure Schmill confirmed to the student newspaper The Tech.
- The wait list produced 10 admits from 561 students offered a place.
- Deferred early applicants made up 175 of the 1,334 admits, about 2.3% of the 7,486 who were deferred.
- Transfer admission accounted for 32 students from 1,346 applicants in the most recent reporting year.
Read those four lines together and the pattern is hard to miss. Each is a lottery with published odds. Only transfer is something a student can plan for in advance; the rest are outcomes, not strategies.
The aid change that families keep missing
On 20 November 2024, MIT announced that from the following autumn, undergraduates from families earning under $200,000 would pay no tuition, and those from families under $100,000 would pay nothing at all, including housing, dining, fees and an allowance for books. Families between $100,000 and $200,000 pay on a sliding scale capped at roughly $23,970 a year, the cost of everything except tuition.
The money behind it is real. MIT earmarked $167.3 million in need-based aid for undergraduates that year, up about 70% from a decade earlier. The median annual cost paid by an undergraduate receiving aid was $12,938, and 87% of the Class of 2024 graduated without debt. The Institute is one of nine U.S. colleges that neither considers an applicant's ability to pay nor leaves demonstrated need unmet.
What the policy did to application volume is harder to pin down. In the same cycle, Regular Action applications rose almost 10% to 17,229 while Early Action applications fell about 4%. MIT has not claimed the aid change caused that shift, and a single cycle cannot establish it either way. What is documented is the timing: the announcement landed about seven weeks before the January deadline, at the point when families were finalising where to apply.
The class that changed after the Supreme Court ruling
In August 2024, MIT became the first highly selective university to publish the demographics of a class admitted after Students for Fair Admissions v. Harvard. The numbers were stark. Sixteen percent of the Class of 2028 identified as Black, Hispanic, Native American or Pacific Islander, against about 25% across the four previous years. Black enrolment fell from 15% to 5%, Latino enrolment from 16% to 11%, and Asian American enrolment rose from 40% to 47%.
Schmill did not soften the result. “I have no doubt that we left out many well-qualified, well-matched applicants from historically under-represented backgrounds who in the past we would have admitted — and who would have excelled,” he said, in an interview with WBUR and in his own published statement. He pointed at the pipeline: Black and Hispanic students are less likely to attend high schools that teach calculus, physics or computer science, the subjects MIT's requirements quietly assume.
The response was recruitment rather than a change to the bar. MIT expanded school visits, joined the Small Town and Rural Students College Network, and scaled up QuestBridge. Two cycles later, the Class of 2029 enrolled 1,155 students from 49 states and 56 countries, 68% of them from public high schools and 20% the first in their families to attend college. The admit rate barely moved.
The student at the laptop will hear from a dozen colleges between December and April, and the number that frightened her all autumn was never a description of her. Four point six percent is an institution's arithmetic. What sits inside her control is narrower and duller than any hack: four years of hard mathematics taken seriously, a test score that reflects what she can actually do, teachers who know her well enough to write something true, and a list of eight or nine colleges where she would be glad to spend four years.
She needs the published numbers, read out plainly in December, by an adult who isn't flinching.
- financial-aid
- college-admissions
- sat-scores
- mit-acceptance-rate
- early-action

