BackgroundPrevious work has suggested that the number of permanent play facilities in school playgrounds and school-based policies on physical activity can influence physical activity in children. However, few comparable studies have used objective measures of physical activity or have had little adjustment for multiple confounders.MethodsPhysical activity was measured by accelerometry over 5 recess periods and 3 full school days in 441 children from 16 primary schools in Dunedin, New Zealand. The number of permanent play facilities (swing, fort, slide, obstacle course, climbing wall etc) in each school playground was counted on three occasions by three researchers following a standardized protocol. Information on school policies pertaining to physical activity and participation in organized sport was collected by questionnaire.ResultsMeasurement of school playgrounds proved to be reliable (ICC 0.89) and consistent over time. Boys were significantly more active than girls (P < 0.001), but little time overall was spent in moderate-vigorous physical activity (MVPA). Boys engaged in MVPA for 32 (SD 17) minutes each day of which 17 (10) took place at school compared with 23 (14) and 11 (7) minutes respectively in girls. Each additional 10-unit increase in play facilities was associated with 3.2% (95% CI 0.0-6.4%) more total activity and 8.3% (0.8-16.3%) more MVPA during recess. By contrast, school policy score was not associated with physical activity in children.ConclusionThe number of permanent play facilities in school playgrounds is associated with higher physical activity in children, whereas no relationship was observed for school policies relating to physical activity. Increasing the number of permanent play facilities may offer a cost-effective long-term approach to increasing activity levels in children.