The purpose of this study was to determine whether physical activity has a positive relationship with school engagement regardless of the presence or absence of a recess or lunch break before the classroom lesson.Data were collected over three ten-week periods: January–April 2014 (Time 1), October–December 2014 (Time 2), and April–June 2015 (Time 3).A cohort of 2194 adolescents (mean age=13.40years, SD=.73) wore an accelerometer during the hour before a mathematics lesson and completed a questionnaire following the mathematics lesson to assess school engagement in that lesson.Linear mixed models indicated that moderate-intensity activity before a mathematics lesson had a positive linear relationship with cognitive engagement (β=.40, p<.05). Recess breaks before a mathematics lesson had a negative relationship with overall, behavioural, emotional, and cognitive engagement (β=−.18, p<.01, β=−.19, p<.01, β=−.13, p=.03, and β=−.13, p=.04, respectively).Promoting moderate-intensity activity prior to mathematics lessons could improve students’ cognitive engagement. Educators should be aware that students tend to demonstrate the lowest levels of school engagement after recess breaks.