Will more recess help boys in school?

What we looked at

Below we summarize the evidence and identify some of the most rigorous and relevant studies on the effects of recess, highlighting differential impacts by gender where available. Many of the proposed mechanisms for improved behavior and academic outcomes flow through the increased physical activity that kids get from recess, so we look at outcomes across four domains—physical, academic, behavioral, and social. We also consider PE and classroom-based interventions when they illuminate mechanisms that might also be at work during recess.

Key evidence

Below we focus on high-quality randomized controlled trials, quasi-experiments, and meta-analyses. Our references section lists the complete set of literature that we consulted.

Texas Fitness Now quasi‑experiment (2007–11; over 1 million student-grade observations)

In one of the few interventions on middle-school (as opposed to elementary school) students, Packham & Street (2019) evaluated Texas Fitness Now (TFN), a four‑year, $37 million grant that mandated 30 minutes of daily PE for low‑income Texas middle schools. They find that the mandate failed to raise average fitness as measured by cardiovascular endurance, strength, flexibility, or average BMI. However, it did lower the proportion of obese students. Academically, math and reading scores were unchanged. They also report slightly lower attendance and more disciplinary reports, and suggest compulsory PE may deter some adolescents who dislike it. They find no differential effects by gender on physical, behavioral, or academic outcomes. An earlier, less rigorous analysis of TFN did find improvement on some measures of fitness but the benefits were greater for girls, possibly because girls tend to be less physically active outside of school.

Lifestyle of our kids (LOOK) 4‑year specialist‑PE trial, Australia (2005-09; 29 schools, 620 third graders)

From grades 3–5, children taught by PE specialists had smaller gains in body‑fat percentage and faster numeracy growth than peers with regular classroom‑led PE. Reading scores remained unchanged. After four years, follow-up found stronger bones and fewer students with high LDL (“bad”) cholesterol; boys’ LDL-C also fell. Social outcomes were not assessed.

Active Smarter Kids (ASK) cluster‑RCT, Norway (2014-15; 7‑month program, 57 schools)

In a cluster‑RCT of 1,129 Norwegian fifth‑graders, treatment schools added 165 minutes of weekly activity through active lessons, short classroom breaks, and active homework; controls kept their usual practice. Accelerometers showed no difference in total or school‑time activity, suggesting control pupils were already quite active. The authors found no overall change in numeracy, reading, English, or composite scores. However, children in the lowest third of baseline numeracy improved by 0.62 SD relative to controls, suggesting some benefits for the weakest students. Differences by sex were minimal with only a small reading difference favoring boys.

Fit & Vaardig 2-year cluster-RCT, Netherlands (2011-15; 12 schools, 500 students)

The Dutch Fit & Vaardig cluster‑RCT randomly assigned 12 elementary schools to receive either three 20-30-minute physically active math‑and‑language lessons or their usual instruction. Total instructional time stayed constant. After two years, the program improved spelling and math but not reading outcomes. For some outcomes (spelling and math speed), effects were weaker in year one and clearer by year two; general math gains were already evident in year one and persisted. The study did not report intervention effects separately for boys and girls, and social‑emotional metrics were not part of the trial.

Playworks cluster‑RCT (2010-11; 25-29 low‑income U.S. schools in 6 cities)

Playworks provides structured, adult-led recess to students in low-income schools and has been the focus of several evaluations. In one evaluation of 29 schools across 6 U.S. cities, over 1,500 4th and 5th-grade students wore accelerometers during recess. Those in treatment schools had significantly higher physical activity levels and were less likely to be observed sedentary than controls. Teachers also reported better student behavior. In an earlier evaluation, one school year of adults‑facilitated recess cut teacher‑reported bullying and exclusion by 43 percent and led to smoother class transitions relative to controls. However, no sex‑stratified analyses were conducted.

Howie et al., 2023 (13-study systematic review of elementary-age recess)

This review of 13 studies (8 quantitative, 5 qualitative, 11 from the U.S.) links longer recess to small gains in attention and teacher-reported behavior. There was no evidence linking greater recess-time to academic harm. The authors concluded that breaks of ≥20 minutes over multiple blocks may be optimal. This review did not look at physical or social outcomes or report sex‑stratified effects.

CDC, 2010 (50-study survey of school-based physical activity interventions)

Across 251 reported associations between activity and academic outcomes, just over half were positive, almost half null, and 1.5% negative. The authors concluded that “increasing time during the school day for physical activity does not appear to take away from academic performance.” Across seven interventions, six reported at least one positive association with behavior and none reported negative effects. Across six recess-focused studies of academic outcomes, all reported at least one positive association and none negative. Few studies reported gender-disaggregated results, and those that did found no consistent pattern.

Álvarez‑Bueno et al., 2017 (18-study meta-analysis of physical activity interventions)

This meta-analysis finds small positive effects on most academic domains, except language-related skills. The included interventions span PE, active lessons or breaks, and activity during recess or lunch, and the authors do not report sex-stratified results. They also find publication bias for some outcomes and note it cannot be ruled out for others.