This study uses population-based estimates to assess the sensitivity and representativeness of an injury surveillance system using a 1-year population-based approach. Data from the Ottawa Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP) site (Children's Hospital of Eastern Ontario) were compared with those from six expansion sites. The overall sensitivity of CHIRPP was 43% of all treated injuries and 57% of injuries treated at emergency departments. CHIRPP was less likely to be representative for older children and more likely to capture children with more severe injuries. The limitations related to using CHIRPP for representing population-based injury remain fairly stable over time. A one-time population-based sample can provide useful information to add to routinely collected injury surveillance.