Schools are important settings for children to engage in physical activity (PA). Most school PA studies focus on PE, and little information exists on the PA of boys and girls during leisure time and how it is affected by environmental variables. PURPOSE:S: To (a) objectively measure leisure-time PA of boys and girls at school before class, at recess, and during lunch, and (b) to assess environmental factors that may influence it. METHODS: SOPLAY (System for Observing Play and Leisure Activity in Youth) observations were made before school, during recess, and at lunch time in 13 elementary schools in 3 districts in California. All PA areas (N=137; mean=10.5 per school; SD=2.8) at schools were identified, and observations were made over 18 months and included 5 days in each school. During an observation each child was coded as Sedentary, Walking, or Vigorous using specialized counters. Entries were simultaneously made for area characteristics (e.g., accessibility, usability; and presence of supervision, loose equipment, and organized activities). Numbers and % of children in the different PA levels in areas were obtained and a summary score, Energy Expenditure Rate (EER), was calculated using heart rate values. RESULTS: Assessors made 2349 visits to areas and observed a total of 36,995 children. More boys (N=1 9,809) than girls (N=17,146) used areas, and they engaged in more Vigorous Physical Activity (VPA; 29.8% vs. 24.6%; p>.001) and Moderate to Vigorous Physical Activity (MVPA; 66.2 vs. 60.0%; p>.001) and had higher EER (0.095 vs. 0.909 kcal/kg/min; p>.001). Children were more active at recess and lunch than before morning classes. While areas were typically accessible and useable, they provided organized activities infrequently (16.5%), especially before school (10.9%). Boys activity levels were similar in areas with and without organized activities, but girls had more MVPA in areas without organized activities (p<.05). Close supervision was provided in 57% of areas. Children had greater MVPA (p<.001) and EER (p<.05) in unsupervised than in supervised areas. About a third (34.7%) of areas had loose, non-permanent, equipment. Children had greater MVPA (p<.001), VPA (p<.001), and EER (p<.001) in areas with equipment. CONCLUSIONS: Children accrued a substantial amount of voluntary PA during leisure. Similar to previous research, boys were more active than girls. Modifiable environmental conditions, such as the provision of equipment, were associated with more activity. Playground supervisors could be trained to provide more vigorous PA and to prompt and reinforce children's activity engagement. Supported by NIH HL73776