The purpose of this study was to investigate the effect of three schedules of practice on high-level violinists' learning. The contextual interference (CI) effect occurs when two or more tasks are practiced in an interleaved manner, which has been shown to impair initial learning but improve retention. How a musician alternates between tasks during practice can have a great impact on learning. This study was designed to explore how an increasing schedule of CI within a practice session would compare to the traditional blocked and random schedules. Sport research has shown a dynamically adaptive schedule to be advantageous, yet this area is relatively unexplored in music. Twenty advanced violinists at a music school practiced three sets of musical excerpts under three practice schedule conditions: blocked (pieces practiced one at a time), increasing contextual interference (blocked and then random), and interleaved (pieces randomly alternating). Recordings were taken at the stages of sight-reading, acquisition (immediately after practice sessions), and retention (24 hours later). Expert ratings of each musician's recordings revealed no effect at the acquisition stage. However, at the delayed retention test, the blocked condition proved to be most effective. Fifty percent of participants perceived blocked practice to be the most effective.