Since more than 50 years, the 'Critical Power' model provides a robust mathematical and physiological framework to study human exercise capacity and fatigue development. Only recently, this concept has been applied to intermittent exercise, which has sparked a lot of interest from both academia and cycling practice. In essence, the application of the CP model to intermittent exercise allows the prediction of an individual's acute energy balance at any time during exercise. For a broad spectrum of people within the world of sports and clinical practice, the practical relevance of having access to this information is great (e.g., to make strategic decisions or improve pacing strategies during races, or individualize training sessions). Although these field applications are very promising, current predictive models have low accuracy as they do not account well for the specific modalities of different exercise protocols, are not tailored to the individual, and lack fundamental knowledge about the underlying physiological processes. Following five years of intensive lab research, we have gathered a large database of exercise trials in which post-exercise recovery capacity was thoroughly investigated. The present project builds on this unique data set with the goal of developing and validating a ground-breaking evidence-based model that enables real-time monitoring of performance capacity during exercise.