We have an incomplete understanding of how humans process the dynamic structure of speech. The significance of the dynamic temporal structure of speech is manifold: For example, the syllabic rhythm of speech appears to be crucial for speech segmentation, and subsequent comprehension. Dynamic temporal cues contribute to prosodic perception, distinguishing speakers in multi-talker listening situations, and allow for predictions in conversational turn-taking. In order to process speech humans likely rely on predictive processing based on prior knowledge of the dynamic structure of speech as well as its content. Predictive processing may contribute to an individual’s ability to flexibly adapt to demanding listening situations and may particularly involve the recruitment of the motor system. This research area aims to investigate the neural computations humans employ to utilize prior knowledge for speech comprehension.