Where test-optional stands in 2026
Test-optional admissions spread rapidly in 2020 and has been receding since. A number of highly selective universities, among them MIT, Dartmouth, Yale, Brown, Harvard, Cornell and several public flagships, now require scores again, citing evidence that scores predict college performance and help identify strong students from under-resourced schools. Many others remain test-optional, and a few are test-blind. The only safe assumption is that each college’s policy is its own and may change for the next cycle.
When to submit
The usual rule works well: submit if your score is at or above the college’s middle-50% range, especially at or above the midpoint; consider withholding if it is well below the 25th percentile. A score inside the range is roughly neutral and usually worth sending, because a real number is more reassuring to a reader than a blank. The good-score guide lists sample ranges, and each college’s Common Data Set has the exact figures.
Context matters more than the number
Admissions readers see scores in context. A 1350 from a school where few students break 1200 is a strong signal; the same 1350 from a competitive magnet school is average for that cohort. If your school profile shows lower typical scores, a score that looks merely “inside the range” may be helping you more than the table suggests.
Scholarships and honors programs
Test-optional admissions does not always mean test-optional aid. Many merit scholarships and honors colleges still use fixed score thresholds, sometimes combined with GPA. If you are counting on merit aid, find those thresholds before deciding not to test.
A decision checklist
- Check each college’s current policy: required, optional, or blind.
- Compare your score with the middle-50% range and the midpoint.
- Check scholarship and honors thresholds separately.
- If you are close but below, consider one more sitting with superscoring in mind.
Last updated September 2026. Policies change each admissions cycle; confirm on each college’s website.