This study investigated the effects of audio, animation, and spatial ability in a computer-based instructional program for biology. The program presented instructional material via test or audio with lean text and included eight instructional sequences presented either via static illustrations or animations. High school students enrolled in a biology course were blocked by spatial ability and randomly assigned to one of four treatments (Text-Static Illustration, Audio-Static Illustration, Test-Animation, Audio-Animation). The study examined the effects of instructional mode (Text versus Audio), illustration mode (Static Illustration versus Animation), and spatial ability (Low versus High) on practice and posttest achievement, attitude, and time. Results for practice achievement indicated that high spatial ability participants achieved more than low spatial ability participants. Similar results for posttest achievement and spatial ability were not found. Participants in the Static Illustration treatments achieved the same as participants in the Animation treatments on both the practice and posttest. Likewise, participants in the Text treatments achieved the same as participants in the Audio treatments on both the practice and posttest. Findings for time-in-program and time-in-instruction indicated that participants in the Animation treatments took significantly more time than participants in the Static Illustration treatments. No time differences of any type were found for participants in the Text versus Audio treatments. Implications for the design of multimedia instruction and topics for future research are included. (Contains 32 references.) (Author/AEF)