As part of a larger study comparing work-related fatal injury of workers in Australia, New Zealand and the United States, an assessment was made of the similarities and differences between the three countries in identifying fatal incidents as work-related or not. The researchers in each country independently classified 333 brief scenarios, describing a variety of fatal incidents, into one of nine categories related to work: worker, bystander, commuter, volunteer, student, suicide, other, unknown and none of the above. Complete agreement with the classification was moderate (62%), but agreement when classifying scenarios as working, not working or unknown was much higher (full agreement for 80% of scenarios; Kappa = 0.71). Only 5% of scenarios were classified differently by all three countries. Other main findings of the study were that there is variation between countries in the interpretation of what is and what is not work-related, and variation in the amount or type of information required to make a definitive classification. Common circumstances described in the scenarios for which there was some disagreement in classification included domestic violence incidents at work, volunteer workers, business trips, social functions connected to work, hobby farmers and some possible bystander incidents that occurred on farms or on the road. The results suggest that, even without the use of standard definitions, comparisons between the datasets of the countries involved in this study can be made with reasonable confidence. However, they also emphasise the importance of minimising ambiguity in the definitions used, and of understanding the manner in which the definitions are applied, when comparing results between studies.