Background: Physical activity throughout the school day may help reduce the risk of childhood obesity. Semi-structured recess is a unique approach to create more physical activity. Therefore, the purpose of this study was to compare the physical activity and enjoyment of urban school children during traditional unstructured recess and semi-structured recess. Method: Children from the Southwestern US (N = 165) wore a NL-1000 piezoelectric accelerometer during their 15-minute lunch recess. Children participated in both their traditional unstructured (no structure and no equipment) recess and semi-structured (organized games and equipment) recess. An enjoyment scale was completed after both types of recess formats. Results: During unstructured recess, children accumulated 1028 ± 356 steps and 4.59 ± 2.2 minutes of MVPA compared to 1156 ± 434 steps and 5.44 ± 2.76 minutes of moderate to vigorous physical activity (MVPA) during semi-structured recess. Paired sample t-tests revealed that children took significantly more steps (t = −4.98; p < 0.001) and MVPA (t = −5.940; p < 0.001) during semi-structured recess. No significant differences were found for enjoyment (p = 0.847) between recesses. Conclusions: It is important for schools to consider creative, semi-structured recess opportunities to increase step counts and time in MVPA (while maintaining enjoyment) and reinforce recess as an important component of a comprehensive school physical activity program.