The purpose of this study was to investigate the relational changes between music performance anxiety (MPA), self-efficacy, performance quality, and behavioural anxiety in five adolescent piano students over a six-week intervention. Additionally, the study explored the effects of a positive self-review self-modelling intervention on adolescent musicians. Self-report measures, performance evaluations, and behavioural anxiety ratings were used to collect data. Results indicate that the relational changes between MPA, self-efficacy, and performance quality are complex. There were no observed relationships between MPA and self-efficacy or performance, suggesting that MPA can have both debilitative and facilitate effects on these variables. Additionally, there was no relationship between MPA and behavioural anxiety, suggesting that students may appear less anxious than they feel. Finally, the results suggest that self-modelling has individual effects on musicians, meaning that self-modelling can provide teachers with a versatile strategy for reducing MPA, improving performance quality, and/or increasing performance confidence.