Summary
Buff Elementary School is a K–5 school in Madras, Oregon, serving 297 students in Jefferson County School District 509J, a district ranked 130th of 141 in Oregon (8th percentile, 0 stars). Buff itself ranks 401st of 709 Oregon elementary schools in 2025–2026 — the 43rd percentile, 2 stars — placing it in the middle of the state and well above its own district. Its history shows real long-term improvement: Buff sat in the bottom 7–9% of Oregon schools around 2014–2017, climbed to the 54th percentile in 2024–2025, and then slipped to 43rd this year.
In 2025–2026 state testing, 44.7% of Buff students were proficient in English language arts (district 25.6%, state 43.8%), 29.8% in math (district 15.9%, state 32.9%), and 23.5% in science (district 14.3%, state 30.6%). Fourth-grade ELA (50%) and third-grade math (41.9%) are bright spots at or above state levels, while fifth-grade math (20% vs. 32.4% statewide) and fifth-grade science (23.5% vs. 30.8%) are the clearest pressure points. Among nearby K–5 schools in the same district, Buff leads in ELA and science and trails only Metolius Elementary School slightly in math, and it outperforms Madras Elementary School and Warm Springs K-8 Academy by wide margins. Terrebonne Community School (57th percentile, 3 stars) still scores higher in math and science despite larger classes and lower per-student spending, while Buff beats Culver Elementary School in all three subjects.
Attendance is one of Buff's strongest features: 28.3% of students were chronically absent, better than the district's 41.3% and the state's 33.5%, and the lowest among its same-district peers. Buff spends $21,823 per student with a 13.5 student–teacher ratio — smaller classes than Madras, Terrebonne, and Culver. Results for some groups stand out: effectiveness rankings for Hispanic students (71st percentile) and low-income students (69th percentile) sit well above the school's overall 43rd percentile, while female students rank lower, at the 34th. Note that reported free/reduced-lunch figures above 100% are reporting artifacts, and online-school data with suppressed scores should be treated as unknown, not zero.
Thank you for your feedback!