Background: Identifying work-related injuries from hospitalization data is not straightforward. How a work-related injury, either acute or non-acute, is defined will affect injury enumeration, injury burden estimates, and subsequent priority setting for prevention activity. Aim: To examine the effect of different case identification criteria on the number and rate of hospitalized work-related injuries and to identify the type of pathologies that are recorded for work-related admissions that have an external cause code. Method: Identified work-related hospitalization admissions from 1 July 2000 to 30 June 2005 were analysed according to different injury case-selection criteria (ie, a principal diagnosis in the ICD-10-AM range S00-T98 and/or an external cause code in the ICD-10-AM range V01-Y98) and different admission restrictions (ie, exclusion of: transfers and type changes; re-admissions within 28 days; or day-only admissions). The principal diagnosis recorded for work-related admissions that had an external cause code were examined. Results: The incidence of identified work-related hospitalized injuries varied widely, from 289 to 611 per 100 000 workers depending on the identification criteria adopted. The exclusion of day-only stays reduced the identified work-related injury admissions by about one-third for each definitional condition. Where only an external cause code was used to identify an injury, the principal diagnoses codes were predominantly injuries, poisoning, and certain other consequences of external causes (74.6%) and diseases of the musculoskeletal system and connective tissue (15.5%). Conclusions: Case-selection criteria adopted to identify work-related hospitalized injuries should be carefully considered. It is recommended that inclusion of certain musculoskeletal conditions that are likely to arise from repetitive minor trauma over time (ie non-acute injuries) should be considered in calculating the estimate of the burden of all work-related hospitalized injuries.