The aim of this study was the identification of accident scenarios and causes by analysing existing accident reports of recognized agricultural occupational accidents with tractors, self-propelled harvesting machinery and materials handling machinery from 2008 to 2010. As a result of a literature-based evaluation of past accident analyses, the narrative text analysis was chosen as an appropriate method. A narrative analysis of the text fields of accident reports that farmers used to report accidents to insurers was conducted to obtain detailed information about the scenarios and causes of accidents. This narrative analysis of reports was made the first time and yielded first insights for identifying antecedents of accidents and potential opportunities for technical based intervention. A literature and internet search was done to discuss and confirm the findings. The narrative text analysis showed that in more than one third of the accidents with tractors and materials handling machinery the vehicle rolled or tipped over. The most relevant accident scenarios with harvesting machinery were being trapped and falling down. The direct comparison of the analysed machinery categories showed that more than 10% of the accidents in each category were caused by technical faults, slippery or muddy terrain and incorrect or inappropriate operation of the vehicle. Accidents with tractors, harvesting machinery and materials handling machinery showed similarities in terms of causes, circumstances and consequences. Certain technical and communicative measures for accident prevention could be used for all three machinery categories. Nevertheless, some individual solutions for accident prevention, which suit each specific machine type, would be necessary.