Is the SAT a Rigorous Predictor of College Success? What the Evidence Shows

Since 8,000 students took the first SAT in 1926, the format, length, and types of questions have all changed, and the exam will continue to evolve in the years ahead. What hasn’t changed is College Board’s focus on making sure the SAT always delivers an accurate and trustworthy signal of students’ readiness for college and career. That’s why the SAT has earned its place as the most respected and widely accepted college entrance exam.   

But it’s fair to ask: How do we know the SAT remains rigorous as it evolves?  

Here’s what the evidence shows about some of the most common questions we get about the SAT. 

Do students who do well on the SAT do better in college?  
SAT scores improve colleges’ ability to predict students’ first-year success (as measured by first-year GPA) by 22% overall and by 38% for STEM majors over high school GPA alone. Here’s what else the evidence behind the SAT shows: SAT scores are strong predictors of college retention and completion. Scores give colleges meaningful information about student readiness beyond high school GPA alone that many colleges use to inform student support efforts including course placement and advising.  
 
Has the SAT gotten easier? Does it still measure what the SAT has always measured? 
No, moving to a digital format has not made the SAT easier. In 2019, on the “old” paper test, and in 2026, on the “new” digital SAT, fewer than 3% of students scored a 1500 or higher.  

The digital SAT, introduced in 2024, may not look like the test many people remember, but it serves the same essential purpose: It invites students to show what they’ve learned by assessing fundamental reading, writing, and math skills. 

Scores on the digital SAT are strongly related to scores on the now-retired paper-and-pencil SAT and other measures of academic strength, such as high-school GPA, PSAT-related test scores, and AP Exam performance.  
 
What about changes to the SAT reading passages and vocabulary questions? Do those make it easier to earn a higher score? 
No, there is no evidence of this. In fact, College Board research finds that shorter reading passages do not make the SAT easier or change how students score. Two studies examined the issue.  

In the first, an analysis involving nearly 200,000 students found that scores were as consistent for students who moved from the paper SAT (with longer passages) to the digital SAT (with shorter passages) as for students who took the paper test twice. In the second, a controlled study found that, on a test with both short and long versions of the same reading passages, students performed the same on questions tied to the longer passages. 

The SAT's current Reading and Writing section exposes students to a broader range of topics, ideas, rhetorical elements, and authors, sampling more of the range of content they will encounter in college and work. The shorter passages allow students to read closely and engage with the content rather than hurriedly skimming for keywords in longer passages. This format ensures measurement of targeted skills (e.g., comprehension, reasoning, vocabulary) rather than endurance, skimming facility, or memory. 

Some researchers at Johns Hopkins University recently looked into the changes and their potential effects.  

That report, commissioned by Thomas Aquinas College and prepared by JHU’s Institute for Education Policy, compared reading content on three SAT test forms, one each from 1980, 1987, and 2026. The report is only descriptive, and its authors plainly state they are not making a claim that the SAT has become easier.  

In their conclusion, the Johns Hopkins authors write: “The new SAT requires students to revise prose for grammar and rhetorical effectiveness and to interpret information presented in charts and graphs — tasks with no counterpart on the 1980 and 1987 forms, which assessed neither writing nor the reading of quantitative displays... .  These are real demands, and on them the newer test plausibly asks more of the test taker, not less. We therefore do not claim that the SAT as a whole has become easier; our evidence cannot support so broad a statement.” 

In a public statement responding to claims that its report shows that the SAT has become easier, the Johns Hopkins Institute for Education Policy was direct: “The report does not make that claim.” 
 
How do we know the SAT predicts college readiness and success? 
Both independent research and College Board’s own research have come to the same conclusion: the SAT is a strong predictor of students' readiness.  

A 2026 review by the Texas A&M Public Policy Research Institute—which was required by state law—gave the SAT a 4 out of 4 score for measuring readiness for college success. The ACT scored a 3.8. The  Classic Learning Test (CLT) scored 1.4.  
 
SAT’s perfect score on predictive validity was based on a sample of more than 117,000 students in 14 Texas public colleges and universities that linked their SAT scores to success in gateway courses such as English Composition and College Algebra.  

What does predictive validity mean and how does the SAT demonstrate it? 
Predictive validity is a test’s ability to predict future outcomes that the assessment is designed to measure. For the SAT, that outcome is college readiness. SAT scores show a consistent relationship to what happens to students after they enroll: their grades, whether they finish a degree, and their outcomes after graduation. Our research finds that these relationships hold from the first year of college through the fourth year, and that they hold among all students with similar high school GPAs.  

Does the SAT predict college grades? 
Yes, SAT scores have a strong relationship to cumulative college GPA, from the first year through the fourth year. Among students with “A” high school GPAs, average first-year college GPA rises from 2.42 for students scoring 600-790 on the SAT to 3.56 for those scoring in the 1400-1600 range. By the fourth year, cumulative GPA ranges from 3.15 to 3.63 across the same SAT score bands. (Marini et al., 2024)  

College Board research has found that SAT Math scores, in particular, reliably identify students at greatest risk of struggling in first-year college math. Across 57 colleges and universities and more than 130,000 students, students who earned Ds, Fs, or withdrew from first-year college math consistently had lower SAT Math scores than students who succeeded. That pattern held before the pandemic and after it. (Marini et al., 2026) 

Does the SAT predict whether students graduate? 
Yes. SAT scores are strongly related to four- and six-year degree completion.  Four-year graduation rates rise from 10% for students scoring 400-590 on the SAT to 74% for students scoring 1400-1600, across a nationally representative sample of more than 850,000 students. (Westrick et al., 2022) 

Does the SAT predict post-college outcomes? 
Independent research from Harvard’s Opportunity Insights links SAT scores to post-college earnings and employment success, regardless of family income or race. (Chetty et al., 2020) 

Why has the SAT changed? 
Since it was introduced in 1926, the SAT has been redesigned many times, innovating and evolving as education, assessment, and technology practices change and modernize.  In 1994, the SAT allowed calculators for the first time on the Math section. In 2005, analogies stopped appearing on the test. In 2016, the test was aligned to widely taught high school standards. It’s gone from Scantron sheets with fill-in-bubbles to Bluebook, College Board’s digital assessment app introduced in 2024. 

Every question is evaluated to ensure accuracy, alignment to high school standards, and that it measures readiness—not wealth, background, or the prep resources students can afford.   

What doesn’t change on the SAT are the readiness skills that colleges need to see.