[Truncated abstract] The inclusion of a warm-up in the form of prior exercise (PE) is generally advocated as a preparatory strategy of choice to improve sprint performance. Although there is evidence that both increasing muscle temperature and mobilising the cardiorespiratory system prior to exercise contribute largely to the benefit of PE on sprint performance, their relative importance is unknown. Another important question relates to situations where an athlete has to engage in a sprint shortly after one or several earlier sprints. Under these conditions, is engaging in mild exercise also the most effective preparatory strategy to adopt prior to sprinting when performed after a previous sprint(s)? It was the primary aim of this thesis to address these questions. Firstly, we hypothesised that there is a temporal shift in the mechanisms responsible for the effect of PE on power output during a maximal sprint effort, with temperature-dependent mechanisms playing a more important role at the onset of the sprint and mobilisation of the cardiorespiratory system playing a more important role later. To test this hypothesis, we compared the responses of a 30-s sprint to different PE protocols designed to control for either muscle temperature or pre-exercise VO2. ... A group of trained athletes was subjected to four consecutive bouts of 30-s sprint, each separated by 20 min of either active recovery at 40% VO2 peak or passive recovery. Our results show that PP, MP-20 and MP-10 did not fall between the first and last sprints, and were not affected by active recovery. In contrast, we found that MP10 and MP30 decrease significantly between the first and last sprint of the passive recovery trial, but not when active recovery is performed between consecutive sprints. Finally, this study also showed that the fall in mean power associated with repeated 30-s sprints in the passive recovery trial resulted primarily from a fall in early, but not late power output. These findings show that the early and late mean power output of repeated sprints respond differently to active and passive recovery, with the decrease in total mean power with repeated 30-s sprints resulting primarily from a fall in early as opposed to either late power output or peak power, thus highlighting the benefit of active recovery as a favourable preparatory strategy for the performance of repeated sprints of short (<10s) or longer duration (<30s), but not for repeated peak power.