Meningococcal disease is a rare, rapidly progressing condition which may be difficult to diagnose, disproportionally affects children, and has high morbidity and mortality. Accurate incidence estimates are needed to monitor the effectiveness of vaccination and treatment. We used linked notification, hospital, mortality and birth data for all children of an Australian state (2000-2007) to estimate the incidence of meningococcal disease. A total of 595 cases were notified, 684 cases had a hospital diagnosis, and 26 cases died from meningococcal disease. All deaths were notified, but only 68% (466/684) of hospitalized cases. Of non-notified hospitalized cases with more than one clinical admission, most (90%, 103/114) did not have meningococcal disease recorded as their final diagnosis, consistent with initial 'false-positive' hospital meningococcal disease diagnosis. After adjusting for false-positive rates in hospital data, capture-recapture estimation suggested that up to four cases of meningococcal disease may not have been captured in either notification or hospital records. The estimated incidence of meningococcal disease in NSW-born and -resident children aged 0-14 years was 5·1-5·4 cases/100 000 child-years at risk, comparable to international estimates using similar methods, but lower than estimates based on hospital data.