Cognitive models can offer deep insights into how stories are comprehended. Models which follow event segmentation theory (EST) focus on the processing of brief episodes or events within a narrative and the boundaries between events. To test the brain mechanisms proposed by EST to occur at the event boundaries we looked at electroencephalographs (EEG) recorded from 49 participants as they were tasked with both listening to and recalling 9 blocks of ~ 6 minute-long audio clips in one of three conditions: single ordered stories, unrelated events from unrelated stories, or single stories in scrambled order. All stimuli were designed to contain event boundaries spaced at semi-regular intervals. Accuracy during an inference recognition task administered after each block was highest in the single ordered stories condition. Analysis 1 examined the effects of event boundary vs. local semantic context on evoked negativities (N400) related to lexical processing of each word. Effects of condition suggest that narrative structure affected lexical processing, more so than event-level structure and sentence-level semantic context. Analysis 2 Examined changes in alpha (8.5-12.5 Hz) and theta (4-8 Hz) band power of the EEG induced by the onset of the event boundary. Boundary-induced changes in both frequencies were recorded, in all conditions. The largest increases were recorded during the ordered stories over large portions of the scalp. How these findings relate to cognitive mechanisms suggested by event segmentation theory is discussed.