A video amusement system by which one or more viewers influence the course of a motion picture as if each viewer were a participant in a real-life drama or dialog. A speech-recognition unit recognizes a few spoken words such as "yes" and "run" spoken by a viewer at branch points in the movie, thus simulating a dialog between the screen actors and the viewer. The apparatus may read an optical videodisc containing independently addressable video frames, blocks of compressed audio, and/or animated cartoon graphics for the multiple story lines which the movie may take. A record retrieval circuit reads blocks of binary-coded control information comprising a branching structure of digital points specifying the frame sequence for each story line. A dispatcher circuit assembles a schedule of cueing commands specifying precisely which video frames, cartoon frames, and portions of audio are to be presented at which instant of time. A cueing circuit executes these commands by generating precisely timed video and audio signals, so that a motion picture with lip-synchronized sound is presented to the viewer. Recordings of the viewers' names may be inserted into the dialog so that the actors speak to each viewer using the viewer's own name. The apparatus can thus provide each viewer with an illusion of individualized and active participation in the motion picture.