Introduction. Learning to read music is a complex task which becomes more difficult if it involves students who have different levels of music knowledge. The cognitive load theory maintains that acquiring new knowledge or the development of new abilities imposes a cognitive effort which depends on: a) instructional design; b) complexity of the task and c) previous knowledge. Method. In the present study eleven students varying in previous musical studies participated in a comparative assessment of music reading exercises using two different kinds of help: usual music instrument and music notation software. Load Cognitive Theory was used as a conceptual framework to determine which interface would be better (imposing low cognitive load) according to both exercises¿ difficulty levels and students´ previous musical studies. The study involved a mixed factorial experimental design with the following independent within-subjects factors and levels 1) Type of Interface: music instrument and music notation software; 2) Music exercises difficulty: level I and level II. The between-subject factor in this study was: 3) Previous Musical Studies: academic and non-academic music studies. Results. Results showed that for the most of variables related with cognitive load, music notation software seemed that imposes less additional cognitive load than music instrument when students had no academic music studies and especially when they studied high-level music exercises. In these cases music notation software was a useful help to support music score study. Discussion or Conclusion. The music notation software allows to perceive clearly the relationship between written notation and sound and probably it makes easier the memorization of these pairs. This kind of media could be a useful tool for students with difficulties and for those who didn´t because of the options´ software, which allows music edition, composing, etc.