Class Size and Learning: What the Research Actually Shows

The Tennessee STAR Study: The Gold Standard

The most rigorous evidence on class size comes from Project STAR (Student/Teacher Achievement Ratio), a randomized experiment conducted in Tennessee from 1985 to 1989. Students in kindergarten through 3rd grade were randomly assigned to small classes (13–17 students), regular classes (22–26), or regular classes with a teacher's aide.

Findings:

• Students in small classes outperformed peers in regular classes on standardized reading and math tests by the equivalent of approximately 3–4 months of additional learning
• Effects were largest for low-income students and Black students, who gained the equivalent of 5–8 months of additional learning
• The teacher's aide condition produced no significant academic benefit compared to regular classes
• Effects persisted: students who had been in small classes in early grades were more likely to take college entrance exams, attend college, and earn higher wages as adults (per Chetty et al., 2011 follow-up study)

Where the Evidence Is Weaker

The STAR study is powerful but has limitations. Subsequent research has found more mixed results:

• Class size reductions produce much smaller effects in middle and high school than in elementary school
• California's 1996 class size reduction initiative (capping K-3 at 20 students) found negligible effects on test scores — likely because the rapid hiring required to staff smaller classes brought many inexperienced teachers into classrooms, offsetting the benefit of smaller classes
• International comparisons are complicated: some high-performing countries (Japan, South Korea) have large classes (30–40 students) combined with excellent teacher quality and strong classroom norms

The key nuance: class size matters, but teacher quality matters more. A class of 28 with an outstanding teacher likely produces better outcomes than a class of 18 with a novice teacher. Class size reduction is only beneficial when teacher quality is maintained during the reduction.

National Average Class Sizes: Benchmarks

Note: student-teacher ratios reported in NCES data (and on sites like MySchoolPeek) are school-wide averages. They include specialist teachers, special education staff, and counselors in the denominator — so the actual classroom ratio students experience is typically 3–5 points higher than the reported student-teacher ratio.

Class Size vs. Class Composition

Some research suggests that class composition — specifically the mix of academic preparation levels among students — matters as much as class size. Students in heterogeneous classes with a wide range of abilities may experience the equivalent of larger-class effects even in numerically small classes, because teachers must differentiate across a wider range of needs.

Tracking and ability grouping (putting high-performing students together) is a separate, contested debate — the key point for parents is that "20 students in a class" means different things in a school that groups by ability versus one that mixes all levels.

How to Evaluate Class Size When Researching Schools

1. Look up the school's student-teacher ratio on MySchoolPeek — ratios below 15:1 are excellent; above 20:1 at the elementary level warrants a closer look
2. Ask the school directly about class sizes — specifically classroom sizes in the grade your child would enter, not school-wide averages
3. Ask about specialist-to-student ratios — reading specialists, math interventionists, and counselors affect how well the school can respond to individual students beyond the main classroom
4. Weight class size more heavily for early elementary grades — the K-3 evidence for class size benefits is strongest, so it matters most to scrutinize ratios at those grades

The Bottom Line

Class size reduction is one of the most effective interventions in education research — when done right. It matters most for young students, for low-income and minority students, and when teacher quality is maintained. It matters less for older students, and its effects can be wiped out if reducing class size requires hiring large numbers of inexperienced teachers. When evaluating schools, use student-teacher ratio as one signal among many, not as a standalone indicator of quality.