DescriptionWe often encounter dynamic stimuli accompanied by some sort of narrative in our daily lives. Little is known about the basic characteristics of eye movements while viewing dynamic stimuli and nothing is known about the basic characteristics of eye movements while viewing dynamic stimuli accompanied by a narrative. In Experiment 1 eye movements were recorded while subjects viewed 2 minute video clips with: audio narration, captioned narration, no narration, and both captions and audio. Results show that subjects spent half their time reading the caption when it was the only form of narration present and up to 40% of their time reading the caption even in the presence of redundant audio narration, which consisted of a narrator reading the caption verbatim. Subjects explored the display to a much wider extent in the absence of any narration relative to when narration was provided. There was no effect of viewing condition on global characteristics of eye movements such as sizes and pause durations. In experiment 2 eye movements were recorded while subjects viewed 15s video clips with concurrent audio narration, audio narration played immediately preceding the clip, and no narration. Results replicated two major findings from Experiment 1, namely: the tendency to explore a larger portion of the display in the absence of narration, and no effect of narration on saccade sizes or pause durations. The reading of the captions in the presence of redundant audio shows that subjects’ looking patterns were motivated by a desire to acquire information in the simplest way that would aid in interpreting the video. Both experiments show profound influences of the narration on the spatial distribution of landing positions, contributing to the growing body of literature asserting that there are powerful top-down influences on where people decide to look.