Abstract. Eye tracking (ET) provides various data like gaze position, pupil size, and eye movement events (blinks, fixations, saccades, etc.). These data can reveal users' cognitive/attentional state but also provide a worthwhile input to human-computer interfaces. Recording and processing such data is an issue especially when integrating ET systems within existing environments. The synchronization of events from the simulation environment and physiological measurement devices with ET data is also a concern. In this paper, we reflect upon a seamless integration process. We gather task fulfillment requirements and confront them with technical constraints. Based on this structured task analysis, we present architecture guidelines regarding efficient ET system integration. Finally, we provide a relevant use case of experimental environment where ET systems have been successfully integrated and discuss architecture solutions.