A case control study examined child pedestrian collisions near schools. More walking was not a risk factor for higher collision rates. One way streets, social disadvantage and crossing guards were associated with higher rates. Higher residential land use density was protective for collisions. There were differences in the influence of social disadvantage by location.