There is a dearth of research attempting to quantify the external (physical) and internal (physiological) demands of amateur boxing performance. Therefore, the purpose of this programme of research was to investigate the external demands of amateur boxing performance, and subsequently, develop a sport-specific simulation protocol that could replicate these demands and the accompanying physiological responses while appraising the reliability and validity of the attempt. To achieve this it was necessary initially to identify key offensive and defensive performance indicators and assess the intra- and inter-observer reliability with which such actions could be quantified. Intra-observer reliability was deemed excellent with high agreement (>92%) for all actions identified. Inter-observer reliability was less impressive (>75%), though remained consistently high nevertheless. Subsequently, research utilising this template quantified the offensive and defensive external demands and effectiveness (i.e. frequency of actions deemed successful) according to the independent and interactive influences of contest outcome, weight class and ability using post-contest video analysis. Main effects, two- and three-way interactions were established when appraising the frequency of actions and their outcomes in relation to the independent variables. Whilst the ability of the boxers evidenced the most prominent impact, contest outcome and weight class remained important influences for most actions. Moreover, substantial (CV >30%) within-group variation was evidenced implicating the role of boxer ‘styles’ and strategies in modifying the demands. The offensive and defensive demands were then supplemented with Global Positioning System (GPS) analyses of the boxers’ sport-specific time-displacement movements. Having established the GPS’s reliability and validity for assessing the boxingmovements, it was observed that boxers typically moved a distance of 35.9 m·min-1 at an average speed of 0.6 m·s-1. Such data was amalgamated with the technical demands to produce a boxing-specific simulation protocol that was reflective of the average competitive demand and thus had the potential to be a boxing conditioning and fitness test (BOXFIT). Despite providing the most valid external demand to-date, owing to confounding influences and within-group dispersion, application of the typical external demand was shown to afford only an approximation of the actual demands in all boxers. As such an issue is characteristic of simulation protocols, the BOXFIT was still employed to evaluate the physiological response and appraise the associated reliability and validity. The internal demand was characterised by a high aerobic cardiopulmonary response (peak heart rate > 189 b·min-1; peak VO2 > 55 ml·kg-1·min-1) coupled with a marked indication of anaerobic energy provision (blood lactate = 4.6 ± 1.3 mmol·l-1). The reliability of the physiological responses elicited by BOXFIT performance was generally sufficient to enable the detection of moderate effects (i.e. 0.6 x pooled SD) and practically relevant changes in physiological and physical performance owing to training and nutritional interventions. However, the BOXFIT-induced responses underestimated selected markers of internal load (e.g. Mean heart rate ≈ -4.5%), questioning its validity. Thus, application of the average external demand typically approximated, rather than replicated, the actual physiology of boxing. With modifications, the validity of the external demands and internal response could be improved. The BOXFIT might therefore be used as part of a boxer’s conditioning, providing a sport-specific means of training and offers an ergonomic framework to assess the impact of systematic, intervention-based changes in boxing-specific exercise physiology.