Understanding how the brain reacts to naturalistic stimuli in continuous tasks and in task-free conditions, such as in resting state studies, is vital for understanding individual variability in spatial and temporal dynamics in real-world brain activity. Reconciling multiple data streams at varying timescales and spatial configurations, however, presents unique methodological challenges. In the present study, we used hidden Markov modelling and partial least squares to study individual differences in brain and behaviour data during a music listening task. We found individual listening patterns between brain, behaviour, and music feature data; as well as group-level patterns distinguishing music listening from resting state tasks. We will present the analysis protocol and discuss how it may be extended to research with clinical populations, particularly those with neurodegeneration.