An experiment comparing the pedagogical effectiveness of five different modes of audio narration in motion and still pictures showed only small differences in sixth graders' learning. Ten experimental groups were formed in which the 351 subjects viewed motion pictures and still slides accompanied by supplementary, redundant, directive, question-posing, and nonlinear narration. The posttest-only experimental design specified no control group. Two weeks after the experiment a composite intellectual ability test was administered in hopes of using the scores from it with the data from the experiment to improve predictions of performance for audiovisual materials. No differences in total amount of learning were revealed for the entire group, for either audio or visual variables. Analysis revealed that students in the low mental ability group scored lowest on presentations that posed questions and that high ability students performed best on treatments with supplementary narration and poorest on question-posing treatments. Only on the knowledge portion of the posttest were there significant differences in learning, favoring the directive, redundant, and supplementary audio styles. A qualitative analysis using the semantic differential showed that motion pictures were more positively received than still pictures. (OH)