This paper describes the use of the two sample capture-recapture method to estimate the completeness of ascertainment of injury events in both routine public hospital discharge statistics and in an active injury surveillance system. For all child pedestrian injuries public hospital discharge statistics were 90% complete and the active surveillance system 77% complete. For non-traffic child pedestrian injuries public hospital discharge statistics were 66% complete, the active surveillance system being 79% complete. Capture-recapture methods have the potential to improve the accuracy of traditional childhood injury surveillance systems.