To develop effective interventions to promote children's PA, insight into PA determinants is necessary. The aim of this study is to investigate the association between parental perceived physical environmental characteristics, and children's location-specific physical activity and children's total moderate- to vigorous-intensity physical activity.

Data were collected between December 2011 and May 2013 as part of the Belgian Environmental Physical Activity Study (BEPAS). Children (9–12 years; n=606) were recruited from 18 elementary schools in Ghent (Belgium). One of their parents completed a questionnaire concerning their physical neighbourhood environment and their children's physical activity in specific locations. Children wore an ActiGraph accelerometer for seven consecutive days. Multilevel logistic regression analyses were conducted.

Parental perception of presence of recreation facilities was positively associated (OR=1.618; CI=1.23, 2.12) with physical activity in public recreation spaces. The presence of a garden (OR=4.468; CI=2.23, 8.96) was positively related to PA in the garden. Street connectivity (OR=0.479; CI=0.33, 0.70) was negatively related to neighborhood physical activity, whereas land use mix accessibility (OR=1.704; CI=1.25, 2.33) and crime safety (OR=1.879; CI=1.29, 2.74) were positively related to neighborhood physical activity. No environmental variables were related to overall moderate- to vigorous-intensity physical activity.

The physical environment relates differently to children's physical activity in specific locations. These findings show that the physical environment is mainly related to PA that actually takes place in the neighborhood and is probably unrelated to PA in other contexts. In order to develop effective interventions, it seems promising to further investigate the relation between location-specific physical activity and different neighborhood environmental correlates.