As of October 2026. Every number below is sourced; the list is at the end.
Yes. Better than high school grades do. As of October 2026 the best evidence we have says SAT and ACT scores are four times more predictive of first-year college grades than a high school GPA. That number comes from Opportunity Insights (Friedman, Sacerdote, Staiger and Tine), published in May 2025. It uses the admissions and transcript records of Ivy-Plus colleges.
The comfortable belief
Here is the story most of us have heard. A test score is really a measure of family income. Families with money hire tutors, the score goes up, and it tells a college nothing except who the parents are. Grades, the story goes, measure the kid.
I understand why that lands. It is not what the researchers found when they went looking.
They checked one thing. Do students from less-resourced high schools, low-income families, or underrepresented backgrounds earn better college grades than richer classmates with the same score? If the score were mostly a wealth signal, they would. They did not. The authors report that the scores show "no calibration bias." In plain words: at the same score, both groups did about the same.
The score is not flattering anybody. It is just useful.
The numbers
| Comparison | Difference in first-year college GPA |
|---|---|
| 1600 SAT or 36 ACT vs. 1200 SAT or 25 ACT | 0.43 points |
| 4.0 high school GPA vs. 3.2 high school GPA | 0.1 points |
Same study, same students. A 400-point gap on the SAT moved first-year GPA four times as much as the whole climb from a 3.2 to a perfect 4.0 in high school. The lower-scoring students were also five times more likely to earn at least one grade of C+ or lower.
Students admitted without a score did about as well as students who scored roughly 1300 SAT, or 27 to 28 ACT.
UT Austin found the same shape in its own records. Students who sent a score earned a first-semester GPA 0.86 points higher. That was one reason the requirement came back.
Where the evidence stops
Be careful with this. The study covers Ivy-Plus colleges: the eight Ivy League schools plus Chicago, Duke, MIT and Stanford. Very selective, not typical. It does not tell you what happens at a big public university. And the UC and CSU campuses do not consider scores at all. We keep the college-by-college policy picture here: https://highschoolreach.com/test-scores.
And 0.43 of a grade point is real but small. It is not a verdict on a kid.
What colleges did with it
23 well-known colleges have put a testing requirement back after studying results like these. Harvard, Stanford, MIT, Yale, Caltech, UT Austin and Cornell are among them. Dartmouth turned its requirement back on for applicants to its Class of 2029 — the students who started college in fall 2025, so high school seniors the year before. Two of its own professors, Bruce Sacerdote and Michele Tine, did that research. Dartmouth's faculty study turned up something else you should know. Students at less-resourced high schools were holding back scores near 1400 that would have helped them. A selective college reads a score against the student's own high school.
A strong score at your kid's school is not a strong score nationally.
The move
Pick the exam first. Then build the calendar back from the test date. Give your student 6 to 12 weeks of real work before the sitting. Prepare for the whole SAT or ACT, never one section. Start with a free mock SAT and a free mock ACT. Then you are deciding from a real baseline, not a guess: prepgurus.com/take-the-next-step, or call us at 949.229.5359.
Questions parents ask
Are test scores really a better predictor than my kid's GPA?
For first-year college grades, yes. The May 2025 Opportunity Insights study found scores four times more predictive than high school grades, with a perfect 4.0 predicting only 0.1 points more first-year GPA than a 3.2.
Isn't the SAT just measuring how much money we have?
The researchers tested that directly and found no calibration bias: at the same score, students from less-resourced high schools and low-income families did about as well in college as their wealthier classmates, not better.
If my kid's school sends few students to selective colleges, is a 1400 good enough to submit?
Dartmouth's faculty study found students at less-resourced high schools were withholding scores near 1400 that would have helped them, because a selective college reads a score against the student's own high school.
Written by Nick Standlea, founder of Test Prep Gurus. Updated October 2026.
Sources
- Opportunity Insights, "Standardized Test Scores and Academic Performance at Ivy-Plus Colleges" (Friedman, Sacerdote, Staiger & Tine, May 2025) — https://opportunityinsights.org/paper/test-scores/
- Dartmouth Undergraduate Admissions, testing policy — https://admissions.dartmouth.edu/apply/testing-policy
- Reach, test-score policy and research layer — https://highschoolreach.com/test-scores
- Test Prep Gurus, program and diagnostics — https://www.prepgurus.com/frequently-asked-questions