Introduction: Most of the research on the design of multimedia instructional materials has addressed how to combine words and pictures to produce effective presentations whereas the development of single representations has received less attention. In this study we explored different ways of presenting single representations. Method: In Experiment 1, participants (n = 36) learned about plate tectonics from a multimedia presentation including either realistic or schematic dynamic graphics. After the presentation, participants solved tests requiring structural and conceptual knowledge. In Experiment 2, participants (n = 40) learned about plate tectonics from a multimedia presentation including narrations spoken in either male or female voice. After the presentation, participants solved retention and transfer tests. Results: The results in Experiment 1 showed that participants in the schematic condition out-performed those in the realistic condition in the conceptual test. In Experiment 2 participants receiving narrations in male voice outperformed those in the female voice condition in the two tests. Discussion: The results in Experiment 1 mean that learners learn better from graphics that eliminate extraneous material, as they allow learners to use all cognitive resources in germane processing. Those in Experiment 2 indicate that learners learn better from agents perceived as competent instructors, as this enhances motivation and, thus, active processing. (Contains 2 tables and 1 figure.)