This Ph.D. project aims at studying the perceptual-motor mechanisms implicated in intersection crossing task. The first part of this manuscript explains the human, economic and societal challenges associated with the successfully complete the intersection crossing task. The main results of the former studies that have examined the identification of risks associated with this type of maneuver at an intersection are developed. These former studies mainly focused on high-level processes (e.g., discrete judgment or decision tasks), we have decided to focus our attention on low-level processes (e.g., visual guidance) during this Ph.D. project. For this to happen, the Ecological Approach to Perception and Action seems to be a relevant theoretical framework for studying this complex task. Indeed, it preserves the natural link between information and movement. The second part of this manuscript is devoted to the experiments completed during this research. In order to carry out our three experiments, virtual reality fixed-base driving simulator was used. This work contributes to better understand the perceptual substrate involved in intersection crossing task. Also, this work identifies different markers of control based on information-movement coupling. Firstly, our results revealed that the perceptual substrate underlying judgments of arrival time of a vehicle moving towards an intersection is distinct from the perceptual substrate underlying the active control of one’s own approach to the same intersection. Secondly, the results have also strengthened the assumption that the control of approach and intersection crossing task is based on information-movement coupling. To conclude, the last part proposes perspectives on applied research. This perspectives focus on the designing adaptive driver assistance systems (ADAS) of new generation.