We used frequency-tagging analysis to quantify the neuronal synchronization between the hippocampus and auditory regions, on the one hand, and the temporal structure of words and syllables of an artificial speech stream, on the other hand (A). Intracranial recordings in the hippocampus and auditory cortex (B) show that auditory regions are highly responsive to syllable frequency whereas the hippocampus responds mostly to word frequency (C). These findings provide direct evidence of the involvement of the hippocampus in the speech segmentation process.