In two experiments, the principles of cognitive load theory were applied to the design of alternatives to conventional music instruction hypothesised to facilitate learning. Experiment 1 demonstrated that spatial integration of visual text and musical notation, and dual-modal delivery of auditory text and musical notation, were superior to the spatially separated placement of the same visual materials, demonstrating the split-attention and modality effects respectively. In Experiment 2 there were four conditions differentiated by the presence or absence of musical notation and the simultaneous or successive presentation of auditory music, written explanations, and musical notation. Results indicated that the presence of music notation had no effects, but that the simultaneous presentation of either two or three information sources was superior to successive presentation. These results provide further empirical support for the need to consider cognitive load theory when designing instruction in any domain. (Contains 4 figures and 2 tables.)