The purpose of this study was to examine where and how children choose to play in four Australian pre-school centers with very different outdoor playgrounds. Using a momentary time sampling direct observation instrument, a total of 960 scans were taken of pre-determined target areas (paths, paved expanses, grass, softfall, sand feature, manufactured functional, manufactured constructive and natural) within four playgrounds over a 30-day period. During each scan, we recorded the number of boys and girls observed in each target area as well as the dominant type of play (functional, constructive, symbolic, self-focused, talking). A total of 2361 observations of children occurred across the four centers. The results revealed the children were using the four playgrounds differently. At the diverse and natural Center A, the most popular space was the natural area and the least popular space was the sandpit. At the small, compact and diverse Center B, children were fairly evenly dispersed, with the most popular areas being the softfall and paved expanse. At the hard and barren Center C, almost half the children were found on the pavement, but the sandpits and natural areas were also popular. Finally, at the large, sparse and old Center D, children were fairly evenly dispersed, but most were observed playing on the softfall. Across all centers, irrespective of target area, the dominant play activity was functional play followed by self-focused play. This article discusses these findings and asks important questions about the design of pre-school playgrounds. In doing so, this study has begun to explain the relationship between the design of outdoor play spaces, children's choices of play locations and their play behaviors. (author abstract)