When the pandemic shut down testing centers in the spring of 2020, colleges had a problem. Students couldn’t sit for the SAT or ACT, applications were still due, and schools needed a workaround. The solution was test-optional admissions: a policy that allowed applicants to apply without submitting standardized test scores, with no penalty for omitting them. It was, at the time, a reasonable emergency measure.

Then it stayed. And stayed. And as it did, the original rationale quietly shifted. What began as a logistical accommodation started to feel like a values statement — a signal from elite institutions that the SAT and ACT were, at best, an imperfect measure of a student’s potential and, at worst, an instrument of inequity. By 2022, more than 80% of four-year colleges had adopted some form of test-optional policy. The message many families received, reasonably enough, was that standardized testing no longer mattered.

But it did, and the data’s now proving it.

Harvard, Yale, Dartmouth, Brown, Cornell, Stanford, MIT and Johns Hopkins have all reinstated testing requirements in the last two years. Princeton announced last October that it would follow for the 2027-28 cycle. Prominent state universities, including LSU, Auburn, the University of Alabama, and the public university systems of Florida and Georgia, are phasing in requirements for fall 2027. The number of students submitting test scores rose 11% in the most recent application cycle.

The test-optional “experiment” is over. The question now is what students and families do with that information and how much time they have left to act on it.

‍

The List of Schools Requiring Standardized Tests Again Is Growing

‍

The dominoes began falling in early 2024, when Dartmouth announced it would reinstate its SAT/ACT requirement after an internal faculty study found that test scores remained the strongest predictor of academic success. What followed was less a debate than a cascade.

The reinstatement isn’t happening uniformly. Nine of the 12 members of the Ivy Plus group — the Ivy League schools plus Stanford, MIT, Duke and the University of Chicago — have now reinstated testing requirements. But the approaches vary. Penn built a hardship waiver into its policy from day one, allowing students who face genuine barriers to accessing a test site to submit an explanation in lieu of a score. Yale went test-flexible rather than test-required, accepting AP or IB scores as a substitute for the SAT or ACT. Carnegie Mellon made it program-specific: required for Computer Science, optional for Fine Arts. The schools are not simply flipping a switch. They are returning to testing deliberately — and with more nuance than the original policy ever had.

Columbia remains the one holdout, and it has said its test-optional policy is permanent. Whether that stance survives the pressure of total peer isolation is another question entirely. Test-optional means something very different at a school surrounded by test-required competitors, and applicants should factor that into their strategy.

‍

The Reversal Rationale: Grades Stopped Telling the Story

‍

Test-optional policies didn’t solve the problem they were designed to address. When testing became optional, admissions officers leaned more heavily on soft metrics: expensive extracurriculars, elite summer programs and polished essays. Those things cost money. Families with resources knew how to play that game. Families without them were left guessing.

Meanwhile, transcripts were becoming increasingly difficult to read. In 1966, only 21.8% of incoming college freshmen reported having an A average in high school. By 2024, that number had climbed to 84%, without any corresponding rise in actual academic performance. Admissions officers at elite schools were staring at a sea of perfect GPAs with no way to differentiate them.

A Dartmouth internal study found that high school grades predicted just 9% of the variance in college performance, while SAT scores alone explained 22%. The schools didn’t abandon test-optional out of nostalgia. They abandoned it because the data gave them no choice.

‍

What “Test-Optional” Means Today

‍

Here’s what the label doesn’t tell you: at most competitive universities, “test-optional” has never meant “test-irrelevant.” At Yale, before the school reinstated its requirement, internal data showed that students who didn’t submit scores had a 2% chance of admission. Students who did submit had a 6% chance. The Dean of Admissions later admitted the policy had become misleading, and said plainly that calling it optional while that gap existed wasn’t honest. Yale has since corrected course, but many schools haven’t.

What that means practically: if your student is applying to a test-optional school and sitting on a strong score, submitting it isn’t optional in any meaningful sense. And if they’re not submitting because they don’t have a strong score, that absence sends its own signal. Admissions officers know how to read what isn’t there.

The consequences don’t stop at the application. At schools that don’t require scores, students who skip the test may face placement exams once they arrive, and a weak result can put them in a remedial English or math course before they’ve earned a single credit toward their degree. For a family writing a full-tuition check, that’s a steep price for a policy that was supposed to make things easier.

There’s also the AP factor, which doesn’t get nearly enough attention. AP exams now function as an audit of a student’s transcript. A student who earns an A in AP Chemistry but scores a 2 on the AP exam has told admissions something — whether they meant to or not. Strong performance on standardized assessments reinforces the rest of the application. Weak performance undermines it.

One practical note: the testing landscape itself has expanded. The Classic Learning Test has emerged as a third option, now accepted by roughly 325 colleges and universities, including the entire University of North Carolina system as of February 2026. Families who’ve been thinking about this as an SAT-or-ACT decision should know there’s more room to maneuver than there used to be.

‍

The Argument Worth Having

‍

Before we close the case, there’s a study worth sitting with — because families deserve a complete picture, not a convenient one.

A recent National Bureau of Economic Research working paper, authored by City University of New York researchers, found that high school GPA was the strongest predictor of first-year college GPA, retention and graduation rates — outperforming SAT scores across the board. At the six-year graduation mark, the effect of SAT scores was roughly one-tenth that of high school GPA. The researchers also found that test-optional admission models, relying solely on GPA, saw relatively little loss in predicting college success compared to models that included standardized test scores. That’s legitimate research. It deserves to be taken seriously.

Here’s the catch: the study was conducted at a large public urban university system. That context matters enormously. The families Admissions Angle works with are largely targeting selective and highly selective institutions — a category where the research tells a very different story. Only 30% of college students attend schools with selective admissions to begin with, and only a fraction of those are the schools where this conversation is actually happening.

At certain T20s, the data reverses. Brown University economist John Friedman found that test scores predict academic outcomes at elite colleges at a rate four times greater than GPA, even after accounting for race, gender and income. His research also found something that should give pause to anyone relying on transcripts alone: students with a perfect 4.0 in high school achieve college GPAs that are less than 0.1 points higher than those of students who came in with a 3.2. At schools where every applicant has near-perfect grades, GPA stops being useful. There simply isn’t enough variation in the data to differentiate one student from another.

This brings us to a more fundamental issue. A GPA is not a standardized measurement. A 4.0 earned at a school with stringent grading and a 4.0 earned at a school where an A is the default aren’t the same credential. Standardized tests were built on a simple premise: that a student in rural Mississippi and a student on the Upper East Side should be measured by the same instrument, because the schools they attended are not the same and never have been.

The CUNY researchers acknowledge that GPA may capture qualities that tests don’t measure well: a student’s ability to stay motivated, manage time, and persist through difficulty over the years. That’s worth taking seriously. It’s also an argument for using both measures rather than abandoning one.

The honest answer to “which predicts success better” is that it depends on which schools you’re studying, which students you’re tracking and what you define as success. That’s a less satisfying answer than a headline, but it also happens to be correct.

What hasn’t changed: when 84% of high school seniors report having an A average, a transcript alone can’t hold the weight it once did. A standardized test isn’t a perfect instrument. But at schools where the margins are thin and the stakes are high, it remains one of the few data points an admissions officer can trust.

‍

What This Means If You’re Starting Now

‍

Most students need at least six months to prepare for standardized testing, which is why starting in the spring of sophomore year is the recommended entry point. That timeline isn’t arbitrary. It’s the difference between building a skill and cramming for a crisis.

If your student is in 8th, 9th or 10th grade, none of this should feel like bad news. It should feel like clarity. The admissions landscape is changing, but it’s veered in a direction that rewards exactly what thoughtful families were already doing: treating test prep as part of an academic plan, not an emergency measure.

The test-optional era made a quiet argument to families: that rigor was optional, that elite schools had found better ways to assess potential, that a student who didn’t test well could still compete at the highest level without one. Some students for whom testing genuinely wasn’t the right measure benefited from that flexibility. Most families, though, took a bet that didn’t pay off.

The schools haven’t changed what they’re looking for. What has changed is the language used to describe it, and for a few years, that language was convincing enough to change behavior. It shouldn’t anymore.

‍

‍