The importance of work accidents’ analysis is undeniable in a continuous improvement and risk reduction process in industrial activities. Despite the existence of multiple models of accident analysis, their operationalization is difficult, so that technicians and decision structures are often tempted to make simple causality assignments, usually related to the fundamental attribution error (usually blaming the operator). To test this and other hypotheses, a study was carried out comparing university students and industrial safety technicians, to understand their causal attribution when faced with a series of real work accidents. Furthermore, it was also analyzed the variables that subjects evaluate as more relevant when confronted with those accounts. The results of this work will be presented allowing not only to analyze the type of causal attributions of the two groups, but also to promote a systematic comparison of the variables considered fundamental in accidents’ analysis. Recognition of existing biases in causal attribution and underestimation or overestimation of the importance of certain explanatory factors will be critical in designing much more efficient training processes.