Children's visual attention to, and comprehension of, a television program was measured. A total of 64 children, equally distributed by sex, with 32 in kindergarten and 32 in fifth grade, were randomly assigned to one of four treatment conditions that crossed two levels of content cues with two levels of sound effects. The content cue conditions provided 35 seconds of additional information which indicated that a dream was occurring while the no content cue conditions did not. The content cue conditions were either preceded or not preceded by one-second sound effects. Visual attention was videotaped during each child's individual viewing session. After viewing, each child answered a 22-item multiple-choice recognition test of inferential, central-concrete, and incidental content. Results demonstrated that sound effects increased attentional responsiveness and inferential recognition better than the content cues, particularly for the youngest children who have the greatest difficulty understanding televised stories. (Author/RH)