What Actually Changed in College Admissions Trends This Cycle

A Data-Driven Look at the 2025–2026 Admissions Cycle

by Lindsey Kundel, Editor-in-Chief, InGenius Journals

Every fall, families ask us some version of the same question: is it actually harder to get in this year than last year? The honest answer usually requires more than a headline acceptance rate, which is why this piece exists. Rather than lean on national reporting alone, we pulled our own internal data from the 2025–2026 cycle — six years of it, in most cases — to see what our own applicant pool actually shows, cycle over cycle. What follows is that data, presented as plainly as we can manage, with the caveats that belong alongside it.

A note on scope before we start: everything below reflects our own students’ outcomes, not a national sample. It’s a real, sizable dataset — 539 students in the most recent cycle alone — but it describes a specific population working with structured admissions support, not the applicant pool at large. Keep that in mind as a lens, not a universal claim.

Our Applicant Pool Has Nearly Doubled in Six Years

In the 2020–2021 cycle, we tracked 343 students across our core categories. By 2025–2026, that number had grown to 539 — a 57% increase over six admissions cycles, tracking closely with the national application-volume growth described throughout this book’s companion ebook, Same Student, Every Page. The growth wasn’t evenly distributed across every category we track, either. Students from Canada nearly tripled over the period, from 29 to a peak of 94 before settling at 84 this cycle. Students in the “other countries” category — international students outside our largest tracked markets — nearly quintupled, from 15 to 74. Domestic applicants held roughly steady, moving in a narrower band between 104 and 144 across the full six years. The clearest single story in this table is international growth, concentrated specifically outside our historically largest sending markets.

Early Round Applications Are Becoming the Default, Not the Exception

Six years ago, roughly 28% of our students applied through a restrictive early round — Restrictive Early Action, Early Decision, Early Decision II, or the MIT/Caltech early programs. That figure dipped to 21% the following cycle, then climbed steadily: 30.8%, 31.5%, 28.5%, and, in the most recent cycle, 32.1% — just under a third of our entire student population applying through some form of binding or restrictive early commitment. That’s a meaningful shift in how families are approaching strategy, and it lines up with something covered in this book’s companion volume: as regular-decision admit rates keep compressing, the statistical advantage of an early round becomes harder to leave on the table, even for students who’d prefer more time to decide.

Outcomes, Cycle by Cycle

Here’s the six-year trend across our headline outcome categories, worth reading as a full picture rather than any single year in isolation:

A few things stand out here. Top 5 offers have grown almost every single cycle, more than tripling from 33 to 61 over six years, even as the underlying applicant pool “only” grew by 57% — meaning growth in our most competitive outcomes has outpaced growth in overall student volume. The share of students landing at least one Top 30 offer has held remarkably steady, hovering in a narrow band between roughly 53% and 59% across every single cycle we tracked, regardless of how the broader admissions landscape shifted year to year. That consistency, frankly, is the more interesting finding than any single year’s peak. It suggests something closer to a stable floor of outcomes for students who go through a structured process, even as the outside environment — record applicant pools, falling admit rates — kept getting less forgiving.

One more figure worth isolating: the share of students who land at least one “reach” offer — meaning admission to a school outside their expected competitive range — has consistently run higher among non-domestic students than domestic ones. In the most recent cycle, 72.7% of non-domestic students achieved at least one reach offer, against 60.5% of domestic students; in the cycle before that, the gap was even wider, 73.3% versus 54.9%. We don’t have a single confirmed explanation for that gap, and we’d rather say so plainly than offer a tidy story that outruns the data. It may reflect differences in how international families build school lists, timing differences in when international students begin working with us, or something else entirely. It’s a real, repeated pattern across multiple cycles, and it’s worth families discussing directly with a counselor rather than us speculating further here.

Where Students Land, by Competitive Tier

Comparing our two most recent cycles side by side shows real movement across nearly every tier:

The Top 10 and Top 20 tiers both saw meaningful year-over-year growth — nearly five points at the very top of the market, over seven points in the Top 20 band — while the Top 30 tier held essentially flat and the Top 50 tier grew by nearly seven points as well. Read together, this looks less like uniform improvement and more like students clustering more heavily at both the very top and the broader “safely strong” tier, without much change in the middle.

What a Student’s Starting Position Predicts

We track something internally we call a School List Score, or SLS — a working assessment of how competitive a student’s overall profile is heading into a given cycle, scored on a rough 1-to-30 scale. It’s not a guarantee of outcome, and we want to be direct about that. But the relationship between starting SLS band and eventual outcomes in the most recent cycle is worth showing plainly, because it’s the clearest illustration we have of how much a strong starting position matters — and, just as importantly, how much upside remains even for students who don’t start in the top band:

A few honest takeaways from this table. Students starting in the top band achieved a perfect Top 30-plus admission rate, and nearly two-thirds landed a Top 10 offer — a genuinely strong outcome, though worth noting it’s drawn from a small sample of 25 students, and small samples move around more than large ones. More interesting to us, and more useful for most families reading this, is what happens further down the table: even students starting in the 13–18 band — a range that yielded zero Top 10 offers this cycle — still achieved a 68% rate of landing at least one Top 50-plus-20-LAC offer, which is a genuinely strong outcome by any reasonable standard, even if it’s not the outcome families sometimes assume is the only one worth working toward. The data doesn’t support the idea that a modest starting position closes off strong outcomes. It supports the idea that starting position should shape realistic strategy and school-list construction, which is a different and more useful thing to know.

The Most-Applied-To Schools, and What Actually Happened

Beyond the UC system, covered in its own chapter of the companion ebook, these ten schools drew the most applications from our students this past year, along with our internal admit rate for each:

A few things worth flagging honestly here. These are our own students’ admit rates, not the schools’ published overall rates, and they reflect a population that’s already gone through significant strategic preparation — they shouldn’t be read as a prediction for any individual applicant. That said, the spread is genuinely informative: Northeastern and UIUC posted admit rates well above what a family might assume from reputation alone, while Cornell and Yale, despite drawing over a hundred applications each from our own students, converted at rates in line with their well-known overall selectivity. Michigan, notably, drew the single highest application volume in our data — nearly 200 applications from our own student population alone — while converting at under 20%, a useful reminder that even a genuinely excellent public university with broad appeal can carry real reach-school odds for out-of-state and international applicants specifically.

The Safety School List Keeps Getting Shorter

This book’s companion ebook argues that the traditional “safety school” is functionally disappearing at the very top of the market. Our own current-cycle analysis backs that up directly, but also gives it more texture than a blanket statement can. As of this cycle, we still consider these schools reliable safety options for the large majority of students: UC Merced and UC Riverside, Rutgers–New Brunswick, Penn State–University Park, UMass Amherst, Michigan State, George Washington University, University of Pittsburgh, Indiana University–Bloomington, Syracuse University, and Fordham University.

But “safety” is increasingly conditional on specifics that didn’t used to matter as much. This cycle, we moved Georgia Tech, for out-of-state applicants, into our “high reach” category specifically for students in the 19–22 SLS band — a meaningful downgrade from where it sat previously. We’re now advising students in that same 19–22 band to add an additional genuine safety school beyond Virginia Tech alone, since Virginia Tech by itself no longer reliably fills that role for that specific band. Students in the stronger 23–26 band are being encouraged to add UNC Chapel Hill and UVA to their lists, where those schools weren’t previously a default recommendation at that tier. And within the UC system specifically — worth reading alongside this book’s UC chapter — UC Irvine has become notably difficult for in-state California applicants specifically, even as it remains more accessible for out-of-state and international students, an asymmetry that surprises a lot of families building a list.

Two major-specific cautions from this cycle’s data, stated plainly because we’d rather be direct than vague: we’re advising economics and business students in the 19–22 SLS band that their realistic odds at Top 20 schools are quite low this cycle, and warning engineering-track students in the 23–26 band specifically about Berkeley and UCLA, where engineering admit rates run considerably below each campus’s overall published rate. Neither of these is a reason to abandon an interest — they’re reasons to build a school list that reflects real, current odds rather than a school’s general reputation.

Which Majors Got Harder, and Which Got a Little Easier

Major-specific difficulty shifted unevenly this cycle, and it’s worth breaking out by field.

Computer science remains difficult, and notably, it remains difficult even after we apply our standard internal downward adjustment to reflect that difficulty in a student’s overall school-list strategy. CS-adjacent majors — artificial intelligence, data science, information sciences — are now being treated with the same level of caution, having become similarly impacted rather than serving as an easier “backdoor” into a similar field, which was sometimes true in past cycles.

Biology, medicine, and public health remained about as difficult as the prior cycle, with no change to our internal difficulty adjustment.

Engineering, across every sub-discipline we track, also held at the same difficulty level as the prior cycle — with the UC-specific caution above being the most notable new wrinkle this year rather than a change in overall national difficulty.

Business and economics is the one genuine bright spot in this year’s major-specific data, and worth flagging as a real shift rather than noise. Last cycle’s data suggested students in the 19–23 SLS range had almost no realistic path into Top 20 business or economics programs. This cycle’s data shows real, if still limited, success for students in that same range — a big enough shift that we’re actively continuing to monitor it rather than treating it as confirmed yet. If it holds through another cycle, it would represent a genuine loosening in one of the fields that’s been hardest to break into over the past several years.

What We’re Watching Going Into Next Cycle

A few things are shaping how we’re approaching the year ahead, worth sharing plainly rather than keeping internal. We’re pushing for earlier progress over the summer specifically — students who arrive at the start of senior year with the bulk of their Common App and at least one school’s supplements substantially drafted tend to have measurably more room to write stronger, better-revised material once the fall workload hits, rather than racing the clock on first drafts in October. UChicago’s early-decision-zero option is one concrete lever we’re using more deliberately for eligible students specifically to build in that earlier momentum. We’re also watching Early Decision II usage specifically — only 76 of our students used it this past cycle, down from 86 the cycle before, a small but real decline we’re actively working to reverse, since ED2 remains a genuinely underused tool for well-positioned students who miss an ED1 outcome but still want the statistical benefit of a binding round.

The Throughline

None of this data exists to be read in isolation from the rest of what we’ve built for you. Read alongside Same Student, Every Page, the pattern underneath all of it is the same one: this admissions landscape rewards specificity, strategy, and an early, honest read on where a student actually stands — not generic effort applied evenly across a generic list. The numbers above are what that looks like in practice, across hundreds of real students, cycle after cycle.

If you’d like a read on where your own student’s profile realistically stands heading into this fall, we’d welcome the conversation.