This talk gives an overview of our current research on event segmentation. Event segmentation describes how the continuous stream of information can be divided into meaningful segments. According to event segmentation theory, various subprocesses are involved, such as the formation of a working event model, the use of already established event schemata, and the permanent comparison between the current model and reality in order to be able to update the model if the prediction error is too high. Studying healthy adults with an event segmentation task and simultaneously recording neurophysiological data via EEG, we were able to show that these subprocesses are reflected in different frequency bands, which underpins the mechanistic chain of processes. Furthermore, we investigated the effects of different cognitive control states in healthy adults as well as the effects of manipulations of neurophysiological signal transmission, e.g. by pharmacological manipulation of the catecholaminergic system. In a further part of our research, we also investigated event segmentation in healthy adolescents and compared their segmentation with that of healthy adults. Here we found that adolescents divide the information stream into longer segments, but could be brought to an "adult" level via appropriate instructions. Importantly, we were also able to show that changes in event segmentation are a possible underlying mechanism in ADHD. Adolescent ADHD patients segmented the information stream into longer segments than healthy peers, and in particular showed altered processing of social cues. Thus, event segmentation could serve as a conceptual framework for mechanistic ADHD concepts.