Brazilian offshore drilling operations are carried out in a highly demanding and complex working environment. The geological characteristics of the most productive Brazilian reservoirs, located on the so-called Pre-Salt basin, actually implies for their exploration the overcoming of water depths greater than 1,500 meters and the construction of wells that can reach up to 7,000 meters of length. In such environment, human reliability plays a primordial role regarding safety issues. On the same token, in the field of risk analysis, the consideration of human behavior holds a unique position and its relevance for the consubstantiation of accidents is widely recognized throughout the literature. Nonetheless, a comprehensive way to model it is still an ongoing effort. In this challenge, human error, understood both as a success or a failure probability, provided the first base for the development of techniques capable to infer the outcome towards man machine interaction (MMI). Such stand point gave birth to a series of tools and methodologies, usually called as human reliability analysis (HRA) first generation tools. In an attempt to address the perceived deficiencies of the first-generation tools, the decade of the 1990’s saw the introduction of HRA models based on cognitive assumptions, naturally called HRA second-generation tools. On its turn, the cognitive modeling perspective has demanded the adoption of more sophisticated architectures that, ultimately, have been leading to the dynamically account of the cognitive process. In this context, the present work aims to contribute for this debate analyzing the use of the Dominance, Influence, Steadiness and Compliance (DISC) personality test in the assessment of accidents likelihood on offshore drilling operations in Brazil. By means of actual data collected during the years 2016 and 2017, the paper develops a qualitative discussion of the foreseen outcomes regarding accident proneness considering the working and the under-pressure profiles advocated on DISC.