This article discusses the main substantive issues surrounding performance analysis and considers future directions in this recently formed sub-discipline of sport science. It is argued that it is insufficient to bring together sport biomechanics and notational analysis on the basis that they share a number of commonalities, such as they both aim to enhance performance, they both make extensive use of information and communications technology, and both are concerned with producing valid and reliable data. Rather, it is suggested that the common factor linking sport biomechanics and notational analysis is that they can both be used to measure and describe the same phenomenon (i.e. emergent pattern formation) at different scales of analysis (e.g. intra-limb, inter-limb and torso, and inter-personal). Key concepts from dynamical system theory, such as self-organization and constraints, can then be used to explain stability, variability and transitions among coordinative states. By adopting a constraints-based approach, performance analysis could be effectively opened up to sport scientists from other sub-disciplines of sport science, such as sport physiology and psychology, rather than solely being the preserve of sport biomechanists and notational analysts. To conclude, consideration is given to how a more unified approach, based on the tenets of dynamical systems theory, could impact on the future of performance analysis.