MapRankingsSchoolsResearchWA · K-12 · ACADEMIC ATLAS

Research & findings

Washington's test-score recovery stalled in 2025-26

Enrollment-weighted SBA proficiency moved half a point up in math and half a point down in ELA this year, to 41% and 52%. Seven years after the pandemic, Washington is still 7 points below its 2018-19 math rate and 10 points below in ELA. Only 1 in 5 schools has regained its own pre-pandemic math level, and 1 in 7 its ELA level.

Published September 10, 2026pandemicrecoverytest scoresSBA2025-26
+0.5 / −0.5 pp
Change in Washington's enrollment-weighted SBA math and ELA proficiency rates from 2024-25 to 2025-26

OSPI published the 2025-26 Smarter Balanced results in September, and they describe a system that has stopped recovering. Weighted by school enrollment, 40.6% of Washington students met standard in math, up from 40.1% the year before. In English language arts the rate was 52.2%, down from 52.7%. Both moves are inside the range of year-to-year noise. The pre-pandemic rates were 48.0% and 61.8%.

Washington enrollment-weighted SBA proficiency rate (% meeting standard), 2018-19 through 2025-26
30%40%50%60%2018-192021-222022-232023-242024-252025-2662%54%53%52%53%52%48%38%39%39%40%41%
ELA proficientMath proficient

OSPI Washington State Report Card via data.wa.gov. School-level All-Students rates weighted by 2025-26 enrollment. 2019-20 and 2020-21 are skipped: the SBA was not given in spring 2020 and was replaced by a sampled interim test in 2020-21.

Four years past the 2021-22 trough, math has climbed 2.4 points and ELA has fallen 1.4. At math's pace of about 0.6 points a year, the state would reach its 2018-19 rate around 2038. ELA has now declined or held flat in three of the four post-pandemic years. The one-point uptick in 2024-25 that looked like the start of an ELA recovery reversed.

Half of schools moved up, half moved down

One-year change in combined SBA math + ELA proficiency, Washington schools, 2024-25 → 2025-26
05001000Gaining (>+5 pp)504Treading water (±5 pp)919Slipping (<−5 pp)513

WA K12 Academic Atlas history.json; 1,936 schools with a complete math + ELA series in both years.

The median school's combined math-plus-ELA rate moved 0.1 points. 976 of 1,936 schools improved and 960 slipped. Elementary schools were dead flat at the median, middle schools moved down a tenth of a point, and high schools up half a point. The schools that moved by more than a few points are the subject of a companion piece.

Where the small gains came from

OSPI statewide SBA proficiency by grade, 2024-25 and 2025-26 (% meeting standard)
GradeMath 2024-25Math 2025-26ChangeELA 2024-25ELA 2025-26Change
351.2%51.8%+0.647.8%47.6%−0.2
448.3%48.4%+0.149.4%49.3%−0.1
541.9%42.2%+0.352.9%52.6%−0.3
638.6%38.4%−0.247.7%46.7%−1.0
739.2%39.3%+0.150.4%50.8%+0.4
835.5%36.5%+1.048.5%48.0%−0.5
1030.7%32.0%+1.358.9%58.8%−0.1

OSPI's published statewide rates, which weight by students tested rather than by school enrollment, so the all-grades figures run about half a point above the enrollment-weighted series used elsewhere in this article. Converted to the share of students expected to test who met standard, the definition OSPI used through 2024-25.

The only moves larger than half a point were in math at the top of the grade span: 10th grade rose 1.3 points to 32.0% and 8th grade rose 1.0 to 36.5%. Those are also the two lowest-scoring grades in the state. Sixth-grade ELA fell a full point, the largest decline anywhere in the table. Elementary math, where the earliest recovery would be expected, barely moved.

Still far from 2019

22% / 15%
Share of Washington schools at or above their own 2018-19 proficiency rate in 2025-26: math (408 of 1,852 schools) and ELA (273 of 1,882)

The median school is 6.4 points below its own 2018-19 math rate and 8.7 points below in ELA. Seven years on, the pandemic drop is still the dominant feature of nearly every school's trend line. The statewide line above understates how few schools have fully recovered, because the schools that did recover tend to be large and pull the weighted average up.

The gaps did not move

OSPI statewide SBA proficiency by student group (% meeting standard)
GroupMath 2023-24Math 2024-25Math 2025-26ELA 2023-24ELA 2024-25ELA 2025-26
All students39.7%40.7%41.2%50.3%50.8%50.6%
Low-income24.1%25.0%25.3%34.6%35.0%34.3%
Not low-income55.3%55.9%55.8%66.2%66.3%65.5%
Asian67.4%68.3%67.9%72.5%72.8%71.8%
White46.2%47.2%47.8%57.2%57.8%57.6%
Hispanic / Latino22.4%23.5%23.9%33.6%34.1%34.0%
Black / African American22.0%23.2%23.8%35.0%35.6%35.3%
English learners15.2%14.7%14.0%15.5%14.9%14.0%
Students with disabilities15.0%15.5%15.3%18.7%19.2%18.7%

OSPI statewide rows, same definition as the grade table above.

The gap between low-income students and everyone else is 30.5 points in math and 31.2 in ELA. Three years ago it was 31.2 and 31.6. That is a narrowing of less than half a point a year, and it came from the top edging down as much as from the bottom edging up. Hispanic and Black students each gained about a point and a half in math over three years. English learners are the one group moving clearly in the wrong direction: their ELA rate has fallen from 15.5% to 14.0% in two years, and their math rate from 15.2% to 14.0%.

OSPI changed how it reports SBA results this year. The 2025-26 file gives each school's proficiency as a share of students who took the test, where every earlier year used the share of students who were expected to take it. The two definitions differ by the participation rate, about 4 to 5 points statewide. Every figure in this article uses the earlier definition, converted exactly from the new file, so that 2025-26 can be compared with 2018-19. The OSPI Report Card website shows the new definition and will read a few points higher than the numbers here.

Methodology

Statewide trend: school-level SBA math and ELA percent-meeting-standard (All Students, All Grades) for 2018-19 and 2021-22 through 2025-26 from the WA K12 Academic Atlas history.json, weighted by each school's 2025-26 enrollment as a constant weight across years, over schools with a reported value in that year. The 2025-26 point excludes the 71 school-subject values that OSPI withheld under its new Top/Bottom suppression rule and that the atlas carries forward from 2024-25, so that point is pure 2025-26 data. The 2025-26 file reports percentlevelN_including with students tested as the denominator; the atlas multiplies the level-3-plus-level-4 share by percent_participation (students tested divided by students expected) to recover the of-expected definition used in every prior year. One-year direction buckets use the sum of math and ELA percentage-point changes between 2024-25 and 2025-26 across the 1,936 schools with a complete series in both years. The share of schools at or above 2018-19 compares each school's 2025-26 rate with its own 2018-19 rate, per subject, over schools with both values and no carried-forward figure. Grade-level and student-group tables use OSPI's State-level rows in the 2023-24, 2024-25, and 2025-26 assessment files (data.wa.gov datasets x73g-mrqp, h5d9-vgwi, fusr-6kv7), converted to the of-expected definition where needed.

Sources

  • OSPI Washington State Report Card — SBA Math and ELA, 2018-19 and 2021-22 through 2025-26 (data.wa.gov Socrata; 2025-26 dataset fusr-6kv7, published 2026-09-10)
  • WA K12 Academic Atlas history.json (per-school year-over-year trend file)

About the data

These figures come from the WA K12 Academic Atlas — an interactive map of every K-12 school in Washington, assembled from twelve public data sources. Browse the map · see the rankings. Reporters: see the note on the research index.

All research · ← Seattle's teachers strike, side by side: the union wants inflation plus 6 points, the district offers 8.8% — and special education is the real fight · How the best-elementary and best-middle-school rankings changed with the 2025-26 scores →