When I was in elementary school, we had two recesses every day: 20 minutes in the morning and 20 minutes in the afternoon. Plus, we had a 30-minute lunch period, and as soon as we finished eating we were free to go outside and play until the bell rang.
So I was a little surprised when our kids started elementary school 30 years later and found that there was almost *no* recess. Even at lunchtime, they all lined up and headed straight for the lunchroom. When they finished (after just 20 minutes), they had to march right back to the classroom for more lessons. The only physical activity they got at school was P.E. class, which happened two or three times a week, and usually involved very structured “fitness” activities like jogging or calisthenics. (I had P.E. every day at school, in addition to recess.)
Was I just spoiled as a kid? Or do recesses actually serve an educational purpose? It’s a lot to ask six-year-olds to sit still for an entire school day, but perhaps they’re up to the challenge. And maybe my kids’ experience isn’t actually representative of a national trend to eliminate elementary-school recess. A recent study led by Romina Barros made national headlines, proclaiming that “Recess improves classroom behavior.” We thought we’d take a closer look at the study and see if the mainstream accounts were accurate.
The researchers analyzed a very large data set collected as part of the Early Childhood Longitudinal Study, which follows over 15,000 children who started kindergarten in 1998-1999 (which happens to be the same year Nora started kindergarten). The children were from schools across the country, both public and private, and the group spanned all racial and socioeconomic backgrounds. The data for this study came from 2001-2002, from the children who were now in third grade. About 11,600 children qualified for the study.
The kids’ teachers answered questionnaires about how much recess was offered in their schools, how well-behaved the kids were, class size, and so on. Parents gave family income, education levels, and other demographic information.
The recess data came in the form of total number of recesses and in number of 15-minute intervals of recess. So my school with its 2, 20-minute recesses would have been reported as 30-45 minutes per day, considered “a lot of recess” under this study. My kids’ school would have counted as “none/minimal break.” Thirty percent of students in the study fell into this classification. Only 20 percent had “a lot of recess.” Of the kids with “none/minimal break,” 64 percent had P.E. class twice or less per week.
Does recess time depend on the region of the country you live in? Here’s that data:

As you can see, most kids where I live now, in the South, have little or no recess. In the West, where I went to elementary school, only 11.5 percent of kids have little or no recess. Public school kids were significantly less likely to have recess than private school kids, as were kids from poorer households.
So does recess really cause kids to behave better in class? Despite what the USA Today headline implies, this study cannot show that. What it did show is that having recess is associated with better classroom behavior ratings. The teachers rated their class’s behavior on a scale of 1 to 5:
- Misbehaves very frequently and is almost always difficult to handle
- Misbehaves frequently and is often difficult to handle
- Misbehaves occasionally
- Behaves well
- Behaves exceptionally well
This graph compares the amount of recess to teacher behavior ratings:

So indeed, kids with more recess were rated as behaving significantly better. Although the difference looks small, remember that we’re talking about thousands of kids here. Also, the difference between a 3 (“misbehaves occasionally”) and 4 (“behaves well”) rating is probably a tangible one. Even after controlling for socioeconomic status, race, and location of the school, recess still correlated significantly with good classroom behavior.
But of course, we still can’t say just from correlation data that recess causes better classroom behavior. It might be that schools with poorer-performing students tend to remove recess in order to attempt to get them to improve: that the bad behavior leads to removal of recess, rather than the other way around.
However, there have been experimental studies on smaller groups that varied the time of recess in fourth-grade classes and found that students were more fidgety and less attentive right before recess than they were after recess, so it does seem likely that having less recess time may be responsible for at least some part of the bad classroom behavior these researchers observed.
The researchers add that the No Child Left Behind Act has led some schools to replace recess with classroom instruction, a tactic that may turn out to be counterproductive if recess really does cause better classroom behavior on a broad scale.
Barros, R., Silver, E., & Stein, R. (2009). School Recess and Group Classroom Behavior PEDIATRICS, 123 (2), 431-436 DOI: 10.1542/peds.2007-2825