BACKGROUND We assessed the effectiveness of a simple intervention for increasing children's physical activity, play, perceived competence/social acceptance, and social skills. METHODS A cluster‐randomized controlled trial was conducted, in which schools were the clusters. Twelve Sydney (Australia) primary schools were randomly allocated to intervention or control conditions, with 226 children (5‐7 years old) selected randomly to participate. Data were collected at baseline and after 13 weeks. The intervention consisted of introducing recycled materials without an obvious play purpose into school playgrounds and a risk‐reframing workshop for parents and teachers. RESULTS Children from the intervention schools increased physical activity and reduced sedentary time while control schools decreased physical activity and increased sedentary time. The intervention yielded increases in total accelerometer counts (β = 9350 counts, 95% CI 3490‐1522, p = .002), minutes of moderate/vigorous physical activity (MVPA) (β = 1.8 min, 95% CI 0.52‐3.12, p = .006), and reductions in sedentary time (β = −2.1 min, 95% CI −3.77‐(−0.51), p = .01). Although the changes in time spent in play and nonplay were not statistically different (p = .08) the effect size (d = .27) indicates clinical significance. CONCLUSIONS This intervention was effective for increasing MVPA during recess and demonstrated capacity to improve play opportunities in school playgrounds.