Abstract This study aimed to examine the effects of redundant on-screen text on student learning outcomes (i.e. comprehension, matching, spatial labeling, and diagram reconstruction) when learning from multimedia instruction. An interactive, learner-controlled multimedia material was developed to teach the points of articulation used to describe human speech sounds. Participants included 137 undergraduate students from a large southwestern university in the U.S. who were randomly assigned to one of two treatment conditions: (1) an audio only treatment where audio descriptions of each point of articulation were provided, (2) an audio with text label treatment where audio descriptions of each point of articulation plus redundant text labels were provided. The results showed that having redundant on-screen text with spoken information was helpful for student learning. Overall, results confirm an instance of the reverse redundancy effect when instructional material is complex; redundant on-screen text is short; and learners have control over the pace of instruction.